mirror of
https://github.com/deepseek-ai/deepseek-harness
synced 2026-08-15 21:04:50 +00:00
Merge remote-tracking branch 'origin/master' into dshw/pr-2423
# Conflicts: # packages/client/ui-sidebar/src/client/SidebarRoot.tsx
This commit is contained in:
_Kerman
@@ -2,5 +2,5 @@
|
||||
# side as of the last confirmed-consistent state. Both languages carry equal authority;
|
||||
# after editing either side, bring the other along and re-record with:
|
||||
# pnpm run verify-translation-pairing --write .agents/notes/implemented/architecture/2026-06-11-event-sourced-sessions.md
|
||||
2026-06-11-event-sourced-sessions.md: 1303ee843cd5f9e02eae7a1f9bc51594be4bb407
|
||||
2026-06-11-event-sourced-sessions.zh.md: 4308251ddc4564fc0547230d10671b6595389a1a
|
||||
2026-06-11-event-sourced-sessions.md: b6d17d2db1d9b489f3d4224683b61e62be485014
|
||||
2026-06-11-event-sourced-sessions.zh.md: a78349ea2385b3c9d32870757bb607f15996196d
|
||||
|
||||
@@ -25,4 +25,4 @@ Ordering contract: the loop claims inbox messages before `agent/pre-step`, opens
|
||||
- Replay, trace, and telemetry are structurally guaranteed, not bolted on.
|
||||
- Persistence stays a plugin concern; the in-memory store ships in dsh-session.
|
||||
- The event vocabulary is merge-extensible (plugins add e.g. compaction events); [session persistence](2026-06-14-session-persistence.md) froze its shape once the log became durable.
|
||||
- Derivation cost grows with log length — compaction (dsh-compact) is the intended mitigation, not log mutation.
|
||||
- Derivation cost grows with log length — compaction (dsh-compaction) is the intended mitigation, not log mutation.
|
||||
|
||||
@@ -25,4 +25,4 @@ MVP 要求严格的基于事件的追踪，以及完全可回放的会话（严
|
||||
- 回放、追踪与遥测在结构上得到保证，而非事后附加。
|
||||
- 持久化仍是插件关注点；内存存储随 dsh-session 一起提供。
|
||||
- 事件词汇可通过合并扩展（插件可添加如压缩（compaction）事件）；[会话持久化](2026-06-14-session-persistence.md)在日志具备持久性后固定了其结构。
|
||||
- 派生成本随日志长度增长，压缩（dsh-compact）是预期的缓解手段，而不是改写日志。
|
||||
- 派生成本随日志长度增长，压缩（dsh-compaction）是预期的缓解手段，而不是改写日志。
|
||||
|
||||
@@ -2,5 +2,5 @@
|
||||
# side as of the last confirmed-consistent state. Both languages carry equal authority;
|
||||
# after editing either side, bring the other along and re-record with:
|
||||
# pnpm run verify-translation-pairing --write .agents/notes/implemented/architecture/2026-06-13-capability-seams.md
|
||||
2026-06-13-capability-seams.md: b2dc124f7bf0b56598466e6521e0764af06075ab
|
||||
2026-06-13-capability-seams.zh.md: 79befde57392888dd6a07236379b0155b68d2f01
|
||||
2026-06-13-capability-seams.md: efb000631c6dfe91ab648b102a14a27c30d790b4
|
||||
2026-06-13-capability-seams.zh.md: bf6571dfa5c3a90f856480bc22c6560ed65b3e77
|
||||
|
||||
@@ -8,15 +8,15 @@ English | [中文](2026-06-13-capability-seams.zh.md)
|
||||
|
||||
The harness has swappable capabilities — bash execution today, sandboxed/remote executors and alternative model providers tomorrow. A capability has three concerns that change at different rates and for different reasons: the *contract* (what the capability is), the *implementation* (how it runs), and the *consumer API* (what the model and other plugins program against). Bundling them in one package couples those rates of change — swapping a local executor for a sandboxed one would churn the tool schemas the model sees, even though the model-facing contract never changed.
|
||||
|
||||
This is distinct from "who provides vs. needs a capability at runtime", which Cordis already answers with services + `inject` (a provider registers `ctx.bash`; a consumer declares `inject: ['bash']` and its fiber pends until the service exists). That mechanism is necessary but doesn't dictate package boundaries; this Agent Note does.
|
||||
This is distinct from "who provides vs. needs a capability at runtime", which Cordis already answers with services + `inject` (a provider registers `ctx.shell`; a consumer declares `inject: ['bash']` and its fiber pends until the service exists). That mechanism is necessary but doesn't dictate package boundaries; this Agent Note does.
|
||||
|
||||
## Decision
|
||||
|
||||
A swappable capability has **three roles**:
|
||||
|
||||
1. **Service Definition** — the Cordis `Service` and vocabulary types owning `ctx.<key>` and depending only on the vocabulary the contract needs (e.g. `dsh-bash`: `BashExecutor`, `BashRunResult`, `BashProcess`). A definition may be an abstract class or a concrete registry service; it is never a TypeScript `interface`.
|
||||
1. **Service Definition** — the Cordis `Service` and vocabulary types owning `ctx.<key>` and depending only on the vocabulary the contract needs (e.g. `dsh-shell`: `ShellExecutor`, `ShellRunResult`, `ShellProcess`). A definition may be an abstract class or a concrete registry service; it is never a TypeScript `interface`.
|
||||
2. **Service provider** — a plugin that supplies or registers an implementation (e.g. `dsh-bash-local`: subprocesses, process-group kills, spill-file truncation). Sandboxed and remote providers are sibling packages implementing or registering against the same Service Definition.
|
||||
3. **Consumer** — what the model and plugins program against (e.g. `dsh-tool-bash`: the `bash` schema, with background handles registered into the generic task runtime). Consumers inject the service key and never import provider-specific types.
|
||||
3. **Consumer** — what the model and plugins program against (e.g. `dsh-tool-bash`: the `bash` schema, with background handles registered into the generic job runtime). Consumers inject the service key and never import provider-specific types.
|
||||
|
||||
Service providers and Consumers then evolve independently: a sandboxed executor replaces `dsh-bash-local` without touching a tool schema.
|
||||
|
||||
@@ -24,7 +24,7 @@ Roles normally use separate packages when they evolve independently, but the spl
|
||||
|
||||
## Terminology: "seam" names the trio, not the interface
|
||||
|
||||
A **seam** is the whole capability — the three roles together: a **Service Definition** (the Cordis `Service` that owns `ctx.<key>` and the vocabulary), one or more **Service providers**, and one or more **Consumers**. `packages/bash` is the canonical example — `dsh-bash` / `dsh-bash-local`+`dsh-bash-sandbox` / `dsh-tool-bash`. A package may own multiple roles, but one role alone is not the seam. The term "seam" is reserved for this complete capability; name a constituent by its role, class, service, contract, or extension point. The [glossary](../../../../docs/glossary.md#capability-seam) is the canonical entry.
|
||||
A **seam** is the whole capability — the three roles together: a **Service Definition** (the Cordis `Service` that owns `ctx.<key>` and the vocabulary), one or more **Service providers**, and one or more **Consumers**. `packages/shell` is the canonical example — `dsh-shell` / `dsh-bash-local`+`dsh-bash-sandbox` / `dsh-tool-bash`. A package may own multiple roles, but one role alone is not the seam. The term "seam" is reserved for this complete capability; name a constituent by its role, class, service, contract, or extension point. The [glossary](../../../../docs/glossary.md#capability-seam) is the canonical entry.
|
||||
|
||||
## Alternatives considered
|
||||
|
||||
|
||||
@@ -8,13 +8,13 @@ Status: implemented
|
||||
|
||||
harness 具有可替换的能力：当前是 bash 执行，未来会有沙箱化／远程执行器和替代模型提供方。一项能力涉及三个关注点，它们以不同速率、因不同原因变化：*约定*（这项能力是什么）、*实现*（它如何运行）、*消费方 API*（模型和其他插件面向什么编程）。将三者捆绑在一个包中会耦合这些变化速率——把本地执行器换成沙箱化执行器时，模型看到的工具 schema 也会被搅动，尽管面向模型的约定从未改变。
|
||||
|
||||
这与「谁在运行时提供、谁需要一项能力」是不同的问题，后者 Cordis 已通过服务 + `inject` 解决（提供方注册 `ctx.bash`；消费方声明 `inject: ['bash']`，其 fiber 挂起直到服务存在）。该机制是必要的，但不决定包的边界；本 Agent Note 决定的是包的边界。
|
||||
这与「谁在运行时提供、谁需要一项能力」是不同的问题，后者 Cordis 已通过服务 + `inject` 解决（提供方注册 `ctx.shell`；消费方声明 `inject: ['bash']`，其 fiber 挂起直到服务存在）。该机制是必要的，但不决定包的边界；本 Agent Note 决定的是包的边界。
|
||||
|
||||
## 决策
|
||||
|
||||
一项可替换的能力包含**三个角色**：
|
||||
|
||||
1. **Service Definition**——拥有 `ctx.<key>` 的 Cordis `Service` 和词汇类型，仅依赖约定所需的词汇（例如 `dsh-bash`：`BashExecutor`、`BashRunResult`、`BashProcess`）。Service Definition 可以是抽象类，也可以是具体的注册表服务；绝不是 TypeScript `interface`。
|
||||
1. **Service Definition**——拥有 `ctx.<key>` 的 Cordis `Service` 和词汇类型，仅依赖约定所需的词汇（例如 `dsh-shell`：`ShellExecutor`、`ShellRunResult`、`ShellProcess`）。Service Definition 可以是抽象类，也可以是具体的注册表服务；绝不是 TypeScript `interface`。
|
||||
2. **Service provider**——提供或注册实现的插件（例如 `dsh-bash-local`：子进程、进程组 kill、spill 文件截断）。沙箱化和远程 Service provider 是依据同一 Service Definition 实现或注册的兄弟包。
|
||||
3. **Consumer**——模型和插件编程所面向的内容（例如 `dsh-tool-bash`：`bash` schema，后台句柄注册到通用任务运行时）。Consumer 注入服务键，从不导入 Service provider 特有的类型。
|
||||
|
||||
@@ -24,7 +24,7 @@ Service provider 与 Consumer 由此独立演进：沙箱化执行器替换 `dsh
|
||||
|
||||
## 术语：seam 指三者组合，而非接口
|
||||
|
||||
一个 **seam** 是完整的能力——三个角色合在一起：**Service Definition**（拥有 `ctx.<key>` 和词汇的 Cordis `Service`）、一个或多个 **Service provider**，以及一个或多个 **Consumer**。`packages/bash` 是规范范例——`dsh-bash` / `dsh-bash-local`+`dsh-bash-sandbox` / `dsh-tool-bash`。一个包可以承担多个角色，但单个角色本身不是 seam。「seam」一词严格保留给这种完整能力；命名其中一个组成部分时，应使用其角色、类、服务、约定或扩展点。[术语表](../../../../docs/glossary.md#capability-seam)是规范条目。
|
||||
一个 **seam** 是完整的能力——三个角色合在一起：**Service Definition**（拥有 `ctx.<key>` 和词汇的 Cordis `Service`）、一个或多个 **Service provider**，以及一个或多个 **Consumer**。`packages/shell` 是规范范例——`dsh-shell` / `dsh-bash-local`+`dsh-bash-sandbox` / `dsh-tool-bash`。一个包可以承担多个角色，但单个角色本身不是 seam。「seam」一词严格保留给这种完整能力；命名其中一个组成部分时，应使用其角色、类、服务、约定或扩展点。[术语表](../../../../docs/glossary.md#capability-seam)是规范条目。
|
||||
|
||||
## 曾考虑的替代方案
|
||||
|
||||
|
||||
@@ -2,5 +2,5 @@
|
||||
# side as of the last confirmed-consistent state. Both languages carry equal authority;
|
||||
# after editing either side, bring the other along and re-record with:
|
||||
# pnpm run verify-translation-pairing --write .agents/notes/implemented/architecture/2026-06-14-session-persistence.md
|
||||
2026-06-14-session-persistence.md: 6961b8f3f4274587c2920fe2b2602aff5efe4afc
|
||||
2026-06-14-session-persistence.zh.md: a99b236167d8f3229de083c917684d517b50ab9c
|
||||
2026-06-14-session-persistence.md: 62228bd2f5b25b13880a563818d08f3a2d52d956
|
||||
2026-06-14-session-persistence.zh.md: b10ceebd95d4d05ae3f7ea620183ad8fffd074ee
|
||||
|
||||
@@ -12,7 +12,7 @@ The [event-sourced model](2026-06-11-event-sourced-sessions.md) makes the append
|
||||
|
||||
## Decision
|
||||
|
||||
Persistence is a **capability seam** with an abstract Service Definition ([capability seams](2026-06-13-capability-seams.md), the `dsh-bash` template), not loop or core logic:
|
||||
Persistence is a **capability seam** with an abstract Service Definition ([capability seams](2026-06-13-capability-seams.md), the `dsh-shell` template), not loop or core logic:
|
||||
|
||||
1. **Interface** (`dsh-session-persistence`, `ctx.sessionPersistence`) — an abstract `SessionPersistence` service: `locate`/`create`/`append`/`prepare`/`load`/`inspect`/`readFrom`/`list`/`listSnapshots`. Its persisted unit IS the existing `SessionEvent` (`{ type, seq, time, data }`), reused verbatim — no conversion type.
|
||||
2. **Implementation** (`dsh-session-persistence-jsonl`) — an append-only logical JSONL log per session: a `SessionHeader` line followed by storage records that losslessly represent the contiguous `SessionEvent` stream. Eligible `assistant/chunk` delta runs use packed rows by default; [checksummed Zstandard frames](2026-07-19-zstandard-jsonl-session-logs.md) are the default physical encoding, with raw lines configurable.
|
||||
|
||||
@@ -12,7 +12,7 @@ Status: implemented
|
||||
|
||||
## 决策
|
||||
|
||||
持久化是一个具有抽象 Service Definition 的**能力 seam**（[能力 seam](2026-06-13-capability-seams.md)，`dsh-bash` 模板），而非循环或核心逻辑：
|
||||
持久化是一个具有抽象 Service Definition 的**能力 seam**（[能力 seam](2026-06-13-capability-seams.md)，`dsh-shell` 模板），而非循环或核心逻辑：
|
||||
|
||||
1. **接口**（`dsh-session-persistence`，`ctx.sessionPersistence`）：一个抽象的 `SessionPersistence` 服务，提供 `locate`/`create`/`append`/`prepare`/`load`/`inspect`/`readFrom`/`list`/`listSnapshots`。其持久化单元就是现有的 `SessionEvent`（`{ type, seq, time, data }`），原样复用，无转换类型。
|
||||
2. **实现**（`dsh-session-persistence-jsonl`）：每个会话一个仅追加的逻辑 JSONL 日志：先是一行 `SessionHeader`，随后是无损表示连续 `SessionEvent` 流的存储记录。符合条件的 `assistant/chunk` 增量连续段默认使用打包行；[带校验和的 Zstandard 帧](2026-07-19-zstandard-jsonl-session-logs.md)是默认物理编码，也可通过配置使用原始行。
|
||||
|
||||
@@ -2,5 +2,5 @@
|
||||
# side as of the last confirmed-consistent state. Both languages carry equal authority;
|
||||
# after editing either side, bring the other along and re-record with:
|
||||
# pnpm run verify-translation-pairing --write .agents/notes/implemented/architecture/2026-06-17-filesystem-capability-seam.md
|
||||
2026-06-17-filesystem-capability-seam.md: 63628d592c17f000ae49f9558597bf8059f04942
|
||||
2026-06-17-filesystem-capability-seam.zh.md: d986869bf10f58c8e01873b1d77105accb7bea76
|
||||
2026-06-17-filesystem-capability-seam.md: 953ef0eaafe8f654bf55c3e5c4208560243c3d0b
|
||||
2026-06-17-filesystem-capability-seam.zh.md: 1073226594acbe8c1c596bb4b2dcc959b20c0769
|
||||
|
||||
@@ -6,7 +6,7 @@ English | [中文](2026-06-17-filesystem-capability-seam.zh.md)
|
||||
|
||||
## Problem
|
||||
|
||||
The harness has a concrete `bash` capability seam (`dsh-bash` / `dsh-bash-local` / `dsh-tool-bash`), but filesystem operations were about to land as model-facing tools without an equivalent seam. If `read`, `write`, and `edit` directly used `node:fs`, the model-facing tool package would own filesystem execution policy, local path resolution, atomic write behavior, text decoding, symlink behavior, and edit semantics all at once.
|
||||
The harness has a concrete `bash` capability seam (`dsh-shell` / `dsh-bash-local` / `dsh-tool-bash`), but filesystem operations were about to land as model-facing tools without an equivalent seam. If `read`, `write`, and `edit` directly used `node:fs`, the model-facing tool package would own filesystem execution policy, local path resolution, atomic write behavior, text decoding, symlink behavior, and edit semantics all at once.
|
||||
|
||||
That couples three concerns that change independently:
|
||||
|
||||
@@ -28,7 +28,7 @@ Filesystem access is a first-class capability seam following [the capability-sea
|
||||
|
||||
The Consumer package depends only on the Service Definition package, never on `dsh-fs-local`. A deployment that wants a different backend loads a different provider for `ctx.fs` without changing the tool schemas or model-facing prompt guidance.
|
||||
|
||||
The read-before-write/edit and observed-state policy is a fourth package, `@deepseek-ai/dsh-fs-policy` (`packages/fs/fs-policy`), contributed through the `fs/*` event gate rather than living on `ctx.fs`; a deployment loading `dsh-tool-fs` also loads `dsh-fs-policy` to get read-before-write/edit. This decision established the three-package boundary; the split of policy off the provider base class is decided by [the split-fs-seam Agent Note](../simplification/2026-06-26-fsspec-style-fs-seam.md), and its realization as an event-gate plugin (not a method service) by [the event-gate Agent Note](2026-06-26-file-context-as-event-gate.md).
|
||||
The read-before-write/edit and observed-state policy is a fourth package, `@deepseek-ai/dsh-fs-observation-policy` (`packages/fs/fs-observation-policy`), contributed through the `fs/*` event gate rather than living on `ctx.fs`; a deployment loading `dsh-tool-fs` also loads `dsh-fs-observation-policy` to get read-before-write/edit. This decision established the three-package boundary; the split of policy off the provider base class is decided by [the split-fs-seam Agent Note](../simplification/2026-06-26-fsspec-style-fs-seam.md), and its realization as an event-gate plugin (not a method service) by [the event-gate Agent Note](2026-06-26-file-context-as-event-gate.md).
|
||||
|
||||
The first backend is deliberately local-only: `dsh-fs-local` implements `ctx.fs` against the host filesystem. Future sibling backends can provide sandboxed, remote, virtual, or project-scoped filesystems behind the same interface.
|
||||
|
||||
@@ -36,7 +36,7 @@ The first consumer is deliberately text-file-only: `dsh-tool-fs` exposes model-f
|
||||
|
||||
Filesystem permissions and sandboxing are not implied by this split. The local backend resolves relative paths from its configured base directory, but containment policy is a separate decision: either a stricter `ctx.fs` implementation enforces it, or a permission/sandbox plugin wraps `tools/execute` and vetoes calls before they reach the consumer.
|
||||
|
||||
Read-before-write/edit and observed state belong to `dsh-fs-policy`, not `ctx.fs`. Through the `fs/*` event gate, the policy records versions per opaque actor and supplies optional mutation expectations; the provider enforces freshness atomically. `dsh-tool-fs` emits the events without depending on the policy. See the [split-seam](../simplification/2026-06-26-fsspec-style-fs-seam.md) and [event-gate](2026-06-26-file-context-as-event-gate.md) Agent Notes.
|
||||
Read-before-write/edit and observed state belong to `dsh-fs-observation-policy`, not `ctx.fs`. Through the `fs/*` event gate, the policy records versions per opaque actor and supplies optional mutation expectations; the provider enforces freshness atomically. `dsh-tool-fs` emits the events without depending on the policy. See the [split-seam](../simplification/2026-06-26-fsspec-style-fs-seam.md) and [event-gate](2026-06-26-file-context-as-event-gate.md) Agent Notes.
|
||||
|
||||
## Package topology
|
||||
|
||||
@@ -47,7 +47,7 @@ The filesystem seam uses the same dependency direction as the bash trio:
|
||||
consumer interface implementation
|
||||
```
|
||||
|
||||
`@deepseek-ai/dsh-fs` depends only on `cordis` plus the repo-wide `HarnessError` base from `@deepseek-ai/dsh-llm`. It declares the `ctx.fs` key, the abstract `FileSystem` service, the vocabulary types shared by backends and consumers, the filesystem error vocabulary, and the `fs/*` policy event vocabulary. It carries no observed-state store and no owner-derivation shape; the events pass an opaque `object` actor that the provider never reads, and the `dsh-fs-policy` plugin owns the owner-derivation shape and the observed-state store on top of those events.
|
||||
`@deepseek-ai/dsh-fs` depends only on `cordis` plus the repo-wide `HarnessError` base from `@deepseek-ai/dsh-llm`. It declares the `ctx.fs` key, the abstract `FileSystem` service, the vocabulary types shared by backends and consumers, the filesystem error vocabulary, and the `fs/*` policy event vocabulary. It carries no observed-state store and no owner-derivation shape; the events pass an opaque `object` actor that the provider never reads, and the `dsh-fs-observation-policy` plugin owns the owner-derivation shape and the observed-state store on top of those events.
|
||||
|
||||
`@deepseek-ai/dsh-fs-local` depends on `@deepseek-ai/dsh-fs` and `cordis`. It subclasses `FileSystem`, registers itself as `ctx.fs`, owns local-backend configuration such as the base directory, and contains all direct `node:fs` / `node:path` access. It holds no observed-state store — freshness is a version token the backend mints and the policy plugin records.
|
||||
|
||||
@@ -68,13 +68,13 @@ The interface covers these semantic operations:
|
||||
- Create or replace a UTF-8 text file.
|
||||
- Edit an existing UTF-8 text file by literal replacement.
|
||||
|
||||
The provider contract also carries the freshness hooks that policy builds on — but the observed-state store and owner derivation live in the `dsh-fs-policy` plugin, not on `ctx.fs`:
|
||||
The provider contract also carries the freshness hooks that policy builds on — but the observed-state store and owner derivation live in the `dsh-fs-observation-policy` plugin, not on `ctx.fs`:
|
||||
|
||||
- The backend mints an opaque `version` token per target (in `stat` and in every read/mutation outcome).
|
||||
- `writeText`/`editText` take an OPTIONAL version expectation: omit it for an unconditional bare-provider mutation, or supply it to guard the mutation inside the backend's atomic critical section.
|
||||
- The `dsh-fs-policy` plugin decides that expectation on `fs/write-intent`/`fs/edit-intent` and records observed versions on `fs/observed`, keyed by an owner it derives from the opaque event actor (normally `exec.agent.session`).
|
||||
- The `dsh-fs-observation-policy` plugin decides that expectation on `fs/write-intent`/`fs/edit-intent` and records observed versions on `fs/observed`, keyed by an owner it derives from the opaque event actor (normally `exec.agent.session`).
|
||||
|
||||
Authorization is version freshness, not a full/partial view distinction: any read records the target's version, and a later write/edit is authorized as long as the file is still at that version — so a windowed read of lines 100-150 authorizes an edit of line 120. The observed-state store is a `WeakMap<owner, Map<targetKey, version>>` inside `dsh-fs-policy`; `dsh-fs` holds none of it and treats the actor as opaque. (This decision first modeled a `FileState` cache with `full`/`partial` views on `ctx.fs`; the split-fs-seam and event-gate notes replaced that with the freshness-based policy plugin described here.)
|
||||
Authorization is version freshness, not a full/partial view distinction: any read records the target's version, and a later write/edit is authorized as long as the file is still at that version — so a windowed read of lines 100-150 authorizes an edit of line 120. The observed-state store is a `WeakMap<owner, Map<targetKey, version>>` inside `dsh-fs-observation-policy`; `dsh-fs` holds none of it and treats the actor as opaque. (This decision first modeled a `FileState` cache with `full`/`partial` views on `ctx.fs`; the split-fs-seam and event-gate notes replaced that with the freshness-based policy plugin described here.)
|
||||
|
||||
Path resolution is explicit and allowed to be async. Local resolution may only normalize a path, but sandboxed/remote/project-scoped backends may need I/O to resolve a user-supplied path into a stable target identity.
|
||||
|
||||
@@ -86,11 +86,11 @@ Resolved targets must expose at least three concepts:
|
||||
|
||||
`targetKey` remains opaque even when another capability shares the provider's execution world. Such consumers ask the provider for `processPath(target)`, `fileUrl(target)`, or `contains(parent, child)`; the [portable execution-world decision](2026-07-28-portable-execution-world-consumers.md) owns why these facts sit on the filesystem seam.
|
||||
|
||||
Read and mutation results must include an opaque file `version`. The local backend derives its token from bigint stat metadata (`dev`, `ino`, `size`, `mtimeNs`, and `ctimeNs`) so same-size rewrites and inode replacement invalidate consumers reliably; a remote backend can use a revision id or hash-like token. The `dsh-fs-policy` plugin records versions for stale checks; consumers may display related metadata but must not interpret the version token.
|
||||
Read and mutation results must include an opaque file `version`. The local backend derives its token from bigint stat metadata (`dev`, `ino`, `size`, `mtimeNs`, and `ctimeNs`) so same-size rewrites and inode replacement invalidate consumers reliably; a remote backend can use a revision id or hash-like token. The `dsh-fs-observation-policy` plugin records versions for stale checks; consumers may display related metadata but must not interpret the version token.
|
||||
|
||||
The provider hands back decoded text: `readText` returns a whole regular text file and `streamText` streams the same text semantics for large files or consumer-owned retention limits. Line windowing, byte ceilings, numbered-line rendering, and total-line accounting live in consumers such as `dsh-tool-fs` and `dsh-lsp-local`. The provider owns regular-file checks, UTF-8 decoding, and binary/NUL rejection; it does not know about line windows, protocol limits, or views.
|
||||
The provider hands back decoded text: `readText` returns a whole regular text file and `streamText` streams the same text semantics for large files or consumer-owned retention limits. Line windowing, byte ceilings, numbered-line rendering, and total-line accounting live in consumers such as `dsh-tool-fs` and `dsh-lsp-stdio`. The provider owns regular-file checks, UTF-8 decoding, and binary/NUL rejection; it does not know about line windows, protocol limits, or views.
|
||||
|
||||
Observed-state recording is not on `ctx.fs`: after a successful read the executor emits `fs/observed`, and the `dsh-fs-policy` plugin records `{ version }` for the deriving owner. There is no `full`/`partial` view — a read at any window records the version, and freshness (not view completeness) authorizes a later write/edit.
|
||||
Observed-state recording is not on `ctx.fs`: after a successful read the executor emits `fs/observed`, and the `dsh-fs-observation-policy` plugin records `{ version }` for the deriving owner. There is no `full`/`partial` view — a read at any window records the version, and freshness (not view completeness) authorizes a later write/edit.
|
||||
|
||||
Full-file writes create or replace UTF-8 text files. Backends may create parent directories when that behavior is supported and documented. Existing non-regular targets are rejected. `writeText` takes an optional expectation: `createIfAbsent` creates a missing target and rejects an existing one with `FS_NOT_OBSERVED` (the path the policy uses for an unobserved owner); `replaceIfVersion` replaces only when the target exists at the observed version, else `FS_STALE_VERSION`; omitting the expectation is the unconditional bare-provider create-or-overwrite. The policy plugin chooses which expectation to supply from the owner's observed state.
|
||||
|
||||
@@ -115,19 +115,19 @@ Each tool follows the same execution shape:
|
||||
1. Validate and normalize model arguments.
|
||||
2. Call the appropriate `ctx.fs` operation.
|
||||
3. Format the result as `ContentBlock[]` for the model.
|
||||
4. Let thrown backend/tool errors flow through `ToolRegistry.execute()`, which converts them into `isError` tool results.
|
||||
4. Let thrown backend/tool errors flow through `ToolRuntime.execute()`, which converts them into `isError` tool results.
|
||||
|
||||
The package registers prompt guidance through `ctx.systemPrompt.section(...)` and registers schemas through `ctx.tools.register(...)`. Tool schemas still flow into the normal prompt assembly path via `SystemPrompt.assemble()` and `ToolRegistry.schemas()`; no agent-loop changes are required.
|
||||
The package registers prompt guidance through `ctx.systemPrompt.section(...)` and registers schemas through `ctx.tools.register(...)`. Tool schemas still flow into the normal prompt assembly path via `SystemPrompt.assemble()` and `ToolRuntime.schemas()`; no agent-loop changes are required.
|
||||
|
||||
The tool package keeps model-facing contracts stable when backends change: a local backend and a remote backend may resolve paths differently internally, but the `read` / `write` / `edit` schemas do not change solely because the backend changes.
|
||||
|
||||
The default deployment requires a prior `read` before updating an existing file with `write` or `edit`. `tool-fs` does not implement this by checking whether a tool named `read` ran: it dispatches the `fs/write-intent`/`fs/edit-intent` events (passing the execution context as the opaque actor), and the `dsh-fs-policy` plugin derives the owner, gates on prior observation, and supplies the version expectation. Any windowed read authorizes a later write/edit as long as the file is unchanged. Creating a new file with `write` does not require prior observation.
|
||||
The default deployment requires a prior `read` before updating an existing file with `write` or `edit`. `tool-fs` does not implement this by checking whether a tool named `read` ran: it dispatches the `fs/write-intent`/`fs/edit-intent` events (passing the execution context as the opaque actor), and the `dsh-fs-observation-policy` plugin derives the owner, gates on prior observation, and supplies the version expectation. Any windowed read authorizes a later write/edit as long as the file is unchanged. Creating a new file with `write` does not require prior observation.
|
||||
|
||||
The root plugin registers the full suite by composing the per-tool registration helpers. It injects `fs`, `tools`, and `systemPrompt`.
|
||||
|
||||
## Testing
|
||||
|
||||
Tests follow the package boundary, not only the user-visible tools: the service contract in `dsh-fs`; real filesystem behavior through the `ctx.fs` interface in `dsh-fs-local` (resolution, symlinks, streaming, binary/UTF-8 rejection, unconditional and version-guarded writes, literal-edit semantics, line-ending preservation, structured `FsError` codes); the consumer surface in `dsh-tool-fs` against the real local provider (mock only the model/clock, never the collaborator); and integration through `ctx.tools.execute()` with and without `dsh-fs-policy`, world-verified by reading files back from disk rather than trusting either the canonical value or rendered content. The observed-state/owner-derivation policy is tested in `dsh-fs-policy`, not here.
|
||||
Tests follow the package boundary, not only the user-visible tools: the service contract in `dsh-fs`; real filesystem behavior through the `ctx.fs` interface in `dsh-fs-local` (resolution, symlinks, streaming, binary/UTF-8 rejection, unconditional and version-guarded writes, literal-edit semantics, line-ending preservation, structured `FsError` codes); the consumer surface in `dsh-tool-fs` against the real local provider (mock only the model/clock, never the collaborator); and integration through `ctx.tools.execute()` with and without `dsh-fs-observation-policy`, world-verified by reading files back from disk rather than trusting either the canonical value or rendered content. The observed-state/owner-derivation policy is tested in `dsh-fs-observation-policy`, not here.
|
||||
|
||||
The defensive-pattern classes this repo has been bitten by are pinned directly:
|
||||
|
||||
@@ -152,11 +152,11 @@ The defensive-pattern classes this repo has been bitten by are pinned directly:
|
||||
|
||||
**Edit semantics are race-prone by nature.** Literal edit is a read-modify-write operation; the guard is the backend's atomic mutation critical section plus the optional version expectation, so concurrent edits settle deterministically — one wins, the other gets `FS_STALE_VERSION`.
|
||||
|
||||
**Observed state does not belong on `ctx.fs`.** Recording what an execution context has seen is workflow policy, not raw filesystem I/O. This decision first placed it inside the filesystem seam; the split-fs-seam note then established that a sandboxed/remote backend should not inherit model-facing observation policy, and moved it into the `dsh-fs-policy` plugin. The provider contract keeps only what write/edit safety genuinely needs at the storage layer — a backend-minted version token and an optional version-guarded mutation — while the policy plugin owns owner derivation, observed-state, and read-before-edit gating over the `fs/*` events.
|
||||
**Observed state does not belong on `ctx.fs`.** Recording what an execution context has seen is workflow policy, not raw filesystem I/O. This decision first placed it inside the filesystem seam; the split-fs-seam note then established that a sandboxed/remote backend should not inherit model-facing observation policy, and moved it into the `dsh-fs-observation-policy` plugin. The provider contract keeps only what write/edit safety genuinely needs at the storage layer — a backend-minted version token and an optional version-guarded mutation — while the policy plugin owns owner derivation, observed-state, and read-before-edit gating over the `fs/*` events.
|
||||
|
||||
**The `resolve`-then-operate shape costs an extra round-trip per call.** Each tool may resolve a path to an `FsTarget` and then issue the read/write/edit as a separate `ctx.fs` call. For the local backend this is negligible (resolution is in-memory path normalization), but a remote/sandboxed backend may turn each step into its own request, so a single `read` can become two network round-trips. Backends where the round-trip matters can cache or fold resolution internally while preserving the observable contract.
|
||||
|
||||
**Observed-state persistence is deferred.** Observed state lives in memory (the `WeakMap` inside `dsh-fs-policy`), so a resumed session conservatively requires files to be read again before write/edit until a future session-event or persistence mechanism makes observation replayable.
|
||||
**Observed-state persistence is deferred.** Observed state lives in memory (the `WeakMap` inside `dsh-fs-observation-policy`), so a resumed session conservatively requires files to be read again before write/edit until a future session-event or persistence mechanism makes observation replayable.
|
||||
|
||||
**Error codes become part of the seam.** `FsError` codes make stale-version and observation failures machine-routable through the existing structured error taxonomy. The cost is that `dsh-fs` imports the shared `HarnessError` base from `dsh-llm`; that dependency is intentional and stays limited to the error vocabulary.
|
||||
|
||||
|
||||
@@ -6,7 +6,7 @@ Status: implemented
|
||||
|
||||
## 问题
|
||||
|
||||
harness 已有一个具体的 `bash` 能力 seam（`dsh-bash` / `dsh-bash-local` / `dsh-tool-bash`），但文件系统操作当时即将作为面向模型的工具落地，却没有等价的 seam。如果 `read`、`write` 和 `edit` 直接使用 `node:fs`，面向模型的工具包就会同时承担文件系统执行策略、本地路径解析、原子写入行为、文本解码、符号链接行为和编辑语义。
|
||||
harness 已有一个具体的 `bash` 能力 seam（`dsh-shell` / `dsh-bash-local` / `dsh-tool-bash`），但文件系统操作当时即将作为面向模型的工具落地，却没有等价的 seam。如果 `read`、`write` 和 `edit` 直接使用 `node:fs`，面向模型的工具包就会同时承担文件系统执行策略、本地路径解析、原子写入行为、文本解码、符号链接行为和编辑语义。
|
||||
|
||||
这把三个独立变化的关注点耦合在了一起：
|
||||
|
||||
@@ -28,7 +28,7 @@ harness 已有一个具体的 `bash` 能力 seam（`dsh-bash` / `dsh-bash-local`
|
||||
|
||||
Consumer 包仅依赖 Service Definition 包，从不依赖 `dsh-fs-local`。需要不同后端的部署只需为 `ctx.fs` 加载不同的提供方，无需改动工具 schema 或面向模型的提示词引导。
|
||||
|
||||
读后写/编辑与观测状态策略是第四个包 `@deepseek-ai/dsh-fs-policy`（`packages/fs/fs-policy`），通过 `fs/*` 事件门控贡献，而非挂在 `ctx.fs` 上；加载 `dsh-tool-fs` 的部署同时加载 `dsh-fs-policy` 以获得读后写/编辑能力。本决策确立了由三个包构成的边界；策略从提供方基类拆出的决策由 [拆分文件系统 seam Agent Note](../simplification/2026-06-26-fsspec-style-fs-seam.md) 做出，其以事件门控插件（而非方法服务）实现的方式由 [事件门控 Agent Note](2026-06-26-file-context-as-event-gate.md) 做出。
|
||||
读后写/编辑与观测状态策略是第四个包 `@deepseek-ai/dsh-fs-observation-policy`（`packages/fs/fs-observation-policy`），通过 `fs/*` 事件门控贡献，而非挂在 `ctx.fs` 上；加载 `dsh-tool-fs` 的部署同时加载 `dsh-fs-observation-policy` 以获得读后写/编辑能力。本决策确立了由三个包构成的边界；策略从提供方基类拆出的决策由 [拆分文件系统 seam Agent Note](../simplification/2026-06-26-fsspec-style-fs-seam.md) 做出，其以事件门控插件（而非方法服务）实现的方式由 [事件门控 Agent Note](2026-06-26-file-context-as-event-gate.md) 做出。
|
||||
|
||||
第一个后端有意仅限本地：`dsh-fs-local` 基于宿主文件系统实现 `ctx.fs`。未来的兄弟后端可在同一接口之后提供沙箱、远程、虚拟或项目作用域的文件系统。
|
||||
|
||||
@@ -36,7 +36,7 @@ Consumer 包仅依赖 Service Definition 包，从不依赖 `dsh-fs-local`。需
|
||||
|
||||
文件系统权限和沙箱并非此拆分所隐含。本地后端从其配置的基目录解析相对路径，但路径包含约束策略是独立的决策：要么由更严格的 `ctx.fs` 实现强制执行，要么由权限/沙箱插件包装 `tools/execute` 并在调用到达消费方之前否决。
|
||||
|
||||
读后写/编辑与观测状态属于 `dsh-fs-policy`，而非 `ctx.fs`。通过 `fs/*` 事件门控，策略按不透明 actor 记录版本，并提供可选的变更期望；提供方原子性地强制新鲜度。`dsh-tool-fs` 发出事件但不依赖策略。见[拆分文件系统 seam](../simplification/2026-06-26-fsspec-style-fs-seam.md)和[事件门控插件](2026-06-26-file-context-as-event-gate.md) Agent Note。
|
||||
读后写/编辑与观测状态属于 `dsh-fs-observation-policy`，而非 `ctx.fs`。通过 `fs/*` 事件门控，策略按不透明 actor 记录版本，并提供可选的变更期望；提供方原子性地强制新鲜度。`dsh-tool-fs` 发出事件但不依赖策略。见[拆分文件系统 seam](../simplification/2026-06-26-fsspec-style-fs-seam.md)和[事件门控插件](2026-06-26-file-context-as-event-gate.md) Agent Note。
|
||||
|
||||
## 包拓扑
|
||||
|
||||
@@ -47,7 +47,7 @@ Consumer 包仅依赖 Service Definition 包，从不依赖 `dsh-fs-local`。需
|
||||
consumer interface implementation
|
||||
```
|
||||
|
||||
`@deepseek-ai/dsh-fs` 仅依赖 `cordis` 加上来自 `@deepseek-ai/dsh-llm` 的仓库级 `HarnessError` 基类。它声明 `ctx.fs` 键、抽象 `FileSystem` 服务、后端和消费方共享的词汇类型、文件系统错误词汇，以及 `fs/*` 策略事件词汇。它不持有观测状态存储，也不持有 owner 推导形态；事件传递一个不透明的 `object` actor，提供方从不读取它，`dsh-fs-policy` 插件在这些事件之上拥有 owner 推导形态和观测状态存储。
|
||||
`@deepseek-ai/dsh-fs` 仅依赖 `cordis` 加上来自 `@deepseek-ai/dsh-llm` 的仓库级 `HarnessError` 基类。它声明 `ctx.fs` 键、抽象 `FileSystem` 服务、后端和消费方共享的词汇类型、文件系统错误词汇，以及 `fs/*` 策略事件词汇。它不持有观测状态存储，也不持有 owner 推导形态；事件传递一个不透明的 `object` actor，提供方从不读取它，`dsh-fs-observation-policy` 插件在这些事件之上拥有 owner 推导形态和观测状态存储。
|
||||
|
||||
`@deepseek-ai/dsh-fs-local` 依赖 `@deepseek-ai/dsh-fs` 和 `cordis`。它继承 `FileSystem`，将自身注册为 `ctx.fs`，拥有本地后端配置（如基目录），并包含所有直接的 `node:fs` / `node:path` 访问。它不持有观测状态存储——新鲜度是后端铸造、策略插件记录的版本令牌。
|
||||
|
||||
@@ -68,13 +68,13 @@ Consumer 包仅依赖 Service Definition 包，从不依赖 `dsh-fs-local`。需
|
||||
- 创建或替换一个 UTF-8 文本文件。
|
||||
- 通过字面替换编辑一个已有的 UTF-8 文本文件。
|
||||
|
||||
提供方约定还携带策略所依赖的新鲜度钩子——但观测状态存储和 owner 推导位于 `dsh-fs-policy` 插件中，而非 `ctx.fs` 上：
|
||||
提供方约定还携带策略所依赖的新鲜度钩子——但观测状态存储和 owner 推导位于 `dsh-fs-observation-policy` 插件中，而非 `ctx.fs` 上：
|
||||
|
||||
- 后端为每个目标铸造一个不透明的 `version` 令牌（在 `stat` 以及每次读取/变更结果中）。
|
||||
- `writeText`/`editText` 接受一个可选的版本期望：省略它表示无条件的裸提供方变更；提供它则在后端的原子临界区内守护变更。
|
||||
- `dsh-fs-policy` 插件在 `fs/write-intent`/`fs/edit-intent` 上决定该期望，并在 `fs/observed` 上记录观测版本，以它从不透明事件 actor 推导出的 owner 为键（通常是 `exec.agent.session`）。
|
||||
- `dsh-fs-observation-policy` 插件在 `fs/write-intent`/`fs/edit-intent` 上决定该期望，并在 `fs/observed` 上记录观测版本，以它从不透明事件 actor 推导出的 owner 为键（通常是 `exec.agent.session`）。
|
||||
|
||||
授权基于版本新鲜度，而非完整/部分视图的区分：任何读取都会记录目标的版本，后续的写入/编辑只要文件仍处于该版本就被授权——因此对第 100-150 行的窗口化读取可以授权对第 120 行的编辑。观测状态存储是 `dsh-fs-policy` 内部的 `WeakMap<owner, Map<targetKey, version>>`；`dsh-fs` 不持有任何此类数据，并将 actor 视为不透明。（本决策最初建模了一个带 `full`/`partial` 视图的 `FileState` 缓存放在 `ctx.fs` 上；拆分文件系统 seam 与事件门控两份笔记将其替换为此处描述的基于新鲜度的策略插件。）
|
||||
授权基于版本新鲜度，而非完整/部分视图的区分：任何读取都会记录目标的版本，后续的写入/编辑只要文件仍处于该版本就被授权——因此对第 100-150 行的窗口化读取可以授权对第 120 行的编辑。观测状态存储是 `dsh-fs-observation-policy` 内部的 `WeakMap<owner, Map<targetKey, version>>`；`dsh-fs` 不持有任何此类数据，并将 actor 视为不透明。（本决策最初建模了一个带 `full`/`partial` 视图的 `FileState` 缓存放在 `ctx.fs` 上；拆分文件系统 seam 与事件门控两份笔记将其替换为此处描述的基于新鲜度的策略插件。）
|
||||
|
||||
路径解析是显式的，允许异步。本地解析可能只做路径规范化，但沙箱/远程/项目作用域的后端可能需要 I/O 才能将用户提供的路径解析为稳定的目标标识。
|
||||
|
||||
@@ -86,11 +86,11 @@ Consumer 包仅依赖 Service Definition 包，从不依赖 `dsh-fs-local`。需
|
||||
|
||||
即使另一项能力共享提供方的执行环境，`targetKey` 仍保持不透明。这类消费方通过提供方的 `processPath(target)`、`fileUrl(target)` 或 `contains(parent, child)` 获取所需事实；[可移植执行环境决策](2026-07-28-portable-execution-world-consumers.md)说明这些事实为何属于文件系统 seam。
|
||||
|
||||
读取和变更结果必须包含不透明的文件 `version`。本地后端从 bigint stat 元数据（`dev`、`ino`、`size`、`mtimeNs` 和 `ctimeNs`）派生令牌，因此同大小重写和 inode 替换都会可靠地使消费方失效；远程后端可以使用 revision id 或类似 hash 的令牌。`dsh-fs-policy` 插件记录版本用于陈旧检查；消费方可以展示相关元数据但禁止解释版本令牌。
|
||||
读取和变更结果必须包含不透明的文件 `version`。本地后端从 bigint stat 元数据（`dev`、`ino`、`size`、`mtimeNs` 和 `ctimeNs`）派生令牌，因此同大小重写和 inode 替换都会可靠地使消费方失效；远程后端可以使用 revision id 或类似 hash 的令牌。`dsh-fs-observation-policy` 插件记录版本用于陈旧检查；消费方可以展示相关元数据但禁止解释版本令牌。
|
||||
|
||||
提供方返回已解码的文本：`readText` 返回整个普通文本文件，`streamText` 为大文件或消费方自有的保留上限流式传输相同的文本语义。行窗口化、字节上限、带行号渲染和总行数统计归 `dsh-tool-fs`、`dsh-lsp-local` 等消费方所有。提供方负责普通文件检查、UTF-8 解码和二进制／NUL 拒绝；它不知道行窗口、协议上限或视图。
|
||||
提供方返回已解码的文本：`readText` 返回整个普通文本文件，`streamText` 为大文件或消费方自有的保留上限流式传输相同的文本语义。行窗口化、字节上限、带行号渲染和总行数统计归 `dsh-tool-fs`、`dsh-lsp-stdio` 等消费方所有。提供方负责普通文件检查、UTF-8 解码和二进制／NUL 拒绝；它不知道行窗口、协议上限或视图。
|
||||
|
||||
观测状态记录不在 `ctx.fs` 上：成功读取后，执行器发出 `fs/observed`，`dsh-fs-policy` 插件为推导出的 owner 记录 `{ version }`。没有 `full`/`partial` 视图——任何窗口的读取都记录版本，新鲜度（而非视图完整性）授权后续的写入/编辑。
|
||||
观测状态记录不在 `ctx.fs` 上：成功读取后，执行器发出 `fs/observed`，`dsh-fs-observation-policy` 插件为推导出的 owner 记录 `{ version }`。没有 `full`/`partial` 视图——任何窗口的读取都记录版本，新鲜度（而非视图完整性）授权后续的写入/编辑。
|
||||
|
||||
全文件写入创建或替换 UTF-8 文本文件。后端在支持且有文档说明时可以创建父目录。已有的非常规目标被拒绝。`writeText` 接受一个可选期望：`createIfAbsent` 创建缺失的目标并拒绝已存在的（报 `FS_NOT_OBSERVED`，这是策略为未观测 owner 使用的路径）；`replaceIfVersion` 仅在目标处于观测版本时替换，否则报 `FS_STALE_VERSION`；省略期望则为无条件的裸提供方创建或覆盖。策略插件根据 owner 的观测状态选择提供哪个期望。
|
||||
|
||||
@@ -115,19 +115,19 @@ Consumer 包仅依赖 Service Definition 包，从不依赖 `dsh-fs-local`。需
|
||||
1. 校验并规范化模型参数。
|
||||
2. 调用相应的 `ctx.fs` 操作。
|
||||
3. 将结果格式化为面向模型的 `ContentBlock[]`。
|
||||
4. 让抛出的后端/工具错误流经 `ToolRegistry.execute()`，由其转换为 `isError` 工具结果。
|
||||
4. 让抛出的后端/工具错误流经 `ToolRuntime.execute()`，由其转换为 `isError` 工具结果。
|
||||
|
||||
该包通过 `ctx.systemPrompt.section(...)` 注册提示词引导，通过 `ctx.tools.register(...)` 注册 schema。工具 schema 仍通过 `SystemPrompt.assemble()` 和 `ToolRegistry.schemas()` 流入正常的提示词组装路径；无需改动 agent loop（智能体循环）。
|
||||
该包通过 `ctx.systemPrompt.section(...)` 注册提示词引导，通过 `ctx.tools.register(...)` 注册 schema。工具 schema 仍通过 `SystemPrompt.assemble()` 和 `ToolRuntime.schemas()` 流入正常的提示词组装路径；无需改动 agent loop（智能体循环）。
|
||||
|
||||
工具包在后端变化时保持面向模型的约定稳定：本地后端和远程后端内部可能以不同方式解析路径，但 `read` / `write` / `edit` schema 不会仅因后端变化而改变。
|
||||
|
||||
默认部署要求在用 `write` 或 `edit` 更新已有文件之前先 `read`。`tool-fs` 不通过检查是否运行过名为 `read` 的工具来实现这一点：它分发 `fs/write-intent`/`fs/edit-intent` 事件（将执行上下文作为不透明 actor 传递），`dsh-fs-policy` 插件推导 owner、对先前观测进行门控并提供版本期望。任何窗口化读取都能授权后续的写入/编辑，只要文件未变。用 `write` 创建新文件不要求先前观测。
|
||||
默认部署要求在用 `write` 或 `edit` 更新已有文件之前先 `read`。`tool-fs` 不通过检查是否运行过名为 `read` 的工具来实现这一点：它分发 `fs/write-intent`/`fs/edit-intent` 事件（将执行上下文作为不透明 actor 传递），`dsh-fs-observation-policy` 插件推导 owner、对先前观测进行门控并提供版本期望。任何窗口化读取都能授权后续的写入/编辑，只要文件未变。用 `write` 创建新文件不要求先前观测。
|
||||
|
||||
根插件通过组合各工具的注册辅助函数来注册完整套件。它注入 `fs`、`tools` 和 `systemPrompt`。
|
||||
|
||||
## 测试
|
||||
|
||||
测试遵循包边界，而不仅是用户可见的工具：`dsh-fs` 中的服务约定；`dsh-fs-local` 中通过 `ctx.fs` 接口测试的真实文件系统行为（解析、符号链接、流式传输、二进制/UTF-8 拒绝、无条件和版本守护的写入、字面编辑语义、行尾保留、结构化 `FsError` 错误码）；`dsh-tool-fs` 中基于真实本地提供方的消费方接口（只 mock 模型/时钟，从不 mock 协作者）；以及通过 `ctx.tools.execute()` 在有和没有 `dsh-fs-policy` 的情况下进行集成测试，通过从磁盘回读文件来验证世界状态，既不信任规范值，也不信任渲染内容。观测状态/owner 推导策略在 `dsh-fs-policy` 中测试，不在此处。
|
||||
测试遵循包边界，而不仅是用户可见的工具：`dsh-fs` 中的服务约定；`dsh-fs-local` 中通过 `ctx.fs` 接口测试的真实文件系统行为（解析、符号链接、流式传输、二进制/UTF-8 拒绝、无条件和版本守护的写入、字面编辑语义、行尾保留、结构化 `FsError` 错误码）；`dsh-tool-fs` 中基于真实本地提供方的消费方接口（只 mock 模型/时钟，从不 mock 协作者）；以及通过 `ctx.tools.execute()` 在有和没有 `dsh-fs-observation-policy` 的情况下进行集成测试，通过从磁盘回读文件来验证世界状态，既不信任规范值，也不信任渲染内容。观测状态/owner 推导策略在 `dsh-fs-observation-policy` 中测试，不在此处。
|
||||
|
||||
本仓库曾踩过的防御性模式类别被直接固定：
|
||||
|
||||
@@ -152,11 +152,11 @@ Consumer 包仅依赖 Service Definition 包，从不依赖 `dsh-fs-local`。需
|
||||
|
||||
**编辑语义天然易受竞争影响。** 字面编辑是读-改-写操作；守护手段是后端的原子变更临界区加上可选的版本期望，因此并发编辑确定性地收敛——一个赢，另一个得到 `FS_STALE_VERSION`。
|
||||
|
||||
**观测状态不属于 `ctx.fs`。** 记录执行上下文看到了什么是工作流策略，而非原始文件系统 I/O。本决策最初将其放在文件系统 seam 内部；拆分文件系统 seam 笔记随后确立了沙箱/远程后端不应继承面向模型的观测策略，并将其移入 `dsh-fs-policy` 插件。提供方约定只保留写入/编辑安全在存储层真正需要的东西——后端铸造的版本令牌和可选的版本守护变更——而策略插件拥有 owner 推导、观测状态和基于 `fs/*` 事件的读后编辑门控。
|
||||
**观测状态不属于 `ctx.fs`。** 记录执行上下文看到了什么是工作流策略，而非原始文件系统 I/O。本决策最初将其放在文件系统 seam 内部；拆分文件系统 seam 笔记随后确立了沙箱/远程后端不应继承面向模型的观测策略，并将其移入 `dsh-fs-observation-policy` 插件。提供方约定只保留写入/编辑安全在存储层真正需要的东西——后端铸造的版本令牌和可选的版本守护变更——而策略插件拥有 owner 推导、观测状态和基于 `fs/*` 事件的读后编辑门控。
|
||||
|
||||
**`resolve` 然后操作的形态每次调用多一次往返。** 每个工具可能先将路径解析为 `FsTarget`，再以单独的 `ctx.fs` 调用发起读取/写入/编辑。对本地后端来说这可以忽略（解析是内存中的路径规范化），但远程/沙箱后端可能将每步变成独立请求，使单次 `read` 变为两次网络往返。往返开销重要的后端可以在内部缓存或折叠解析，同时保持可观测约定不变。
|
||||
|
||||
**观测状态持久化被推迟。** 观测状态存在于内存中（`dsh-fs-policy` 内部的 `WeakMap`），因此恢复的会话保守地要求文件在写入/编辑前重新读取，直到未来的会话事件或持久化机制使观测可回放。
|
||||
**观测状态持久化被推迟。** 观测状态存在于内存中（`dsh-fs-observation-policy` 内部的 `WeakMap`），因此恢复的会话保守地要求文件在写入/编辑前重新读取，直到未来的会话事件或持久化机制使观测可回放。
|
||||
|
||||
**错误码成为 seam 的一部分。** `FsError` 错误码使陈旧版本和观测失败可通过既有的结构化错误分类体系进行机器路由。代价是 `dsh-fs` 从 `dsh-llm` 导入共享的 `HarnessError` 基类；该依赖是有意为之且限于错误词汇。
|
||||
|
||||
|
||||
@@ -2,5 +2,5 @@
|
||||
# side as of the last confirmed-consistent state. Both languages carry equal authority;
|
||||
# after editing either side, bring the other along and re-record with:
|
||||
# pnpm run verify-translation-pairing --write .agents/notes/implemented/architecture/2026-06-18-agent-lifecycle-and-ownership-contracts.md
|
||||
2026-06-18-agent-lifecycle-and-ownership-contracts.md: c3522d18dad2703664f12364e4d70cc057a3a945
|
||||
2026-06-18-agent-lifecycle-and-ownership-contracts.zh.md: 16781c8a5d1a1c67d1fe9296b0f34f2a00b11caf
|
||||
2026-06-18-agent-lifecycle-and-ownership-contracts.md: f0f9f90b15dee9155cfb1e8c772503642ff34be3
|
||||
2026-06-18-agent-lifecycle-and-ownership-contracts.zh.md: a5f6ca10dbe7c71281cccdaf83fff95139ad2d79
|
||||
|
||||
@@ -6,7 +6,7 @@ English | [中文](2026-06-18-agent-lifecycle-and-ownership-contracts.zh.md)
|
||||
|
||||
## Problem
|
||||
|
||||
Several ACP and tool-bash limitations were symptoms of the same missing ownership contract: plugins could create or resume agents through `ctx.agents`, but they could not own and dispose one agent independently, and long-running bash tasks carried no stable owner in the executor itself. ACP aborted and awaited agents on disconnect but could not unregister just that session's agent; `session/cancel` could not cancel queued-but-not-yet-started work; and `tool-bash` kept task ownership in a plugin-local `Map`, so an HMR reload could make an old task look unowned.
|
||||
Several ACP and tool-bash limitations were symptoms of the same missing ownership contract: plugins could create or resume agents through `ctx.agents`, but they could not own and dispose one agent independently, and long-running bash tasks carried no stable owner in the executor itself. ACP aborted and awaited agents on disconnect but could not unregister just that session's agent; `session/cancel` could not cancel queued-but-not-yet-started work; and `tool-bash` kept job ownership in a plugin-local `Map`, so an HMR reload could make an old task look unowned.
|
||||
|
||||
## Decision
|
||||
|
||||
@@ -24,7 +24,7 @@ A new `cancel()` verb on the `Agent` interface — the single public stop primit
|
||||
|
||||
### 3. Bash owner token in the Service Definition
|
||||
|
||||
Background-task ownership moved from a `tool-bash` plugin-local `Map<string, Agent>` into the executor. `BashExecRequest` gains an optional `owner?: string`; the resolved `BashExecSpec` carries it as required-but-nullable `owner: string | undefined` (a forgotten owner is a visible `undefined`, never a silently-absent property). The executor stores the token on its task and exposes it via a new `BashExecutor.ownerOf(id): string | undefined` method (NOT on the public `BashTask` — one read path, no redundant API). `tool-bash` deletes its `Map` entirely: it stamps `exec.agent?.id` (the shared registry/session id) as the owner at `start`, and `bash_output`/`bash_kill` compare `ctx.bash.ownerOf(id)` to the caller's token with `!== undefined` semantics (an empty-string token is still a real owner). The completion notice finds the live agent by scanning `ctx.get('agents')?.list()` for `agent.id === ownerToken` (read via `ctx.get` — `onTaskDone` runs on the bash fiber, a foreign fiber, where the `ctx.agents` proxy would throw). Because ownership now lives on the task in the executor (disposed with the `dsh-bash` fiber), it SURVIVES a `tool-bash` HMR reload — closing the old `XXX(tool-bash-owner-hmr)` gap. (The `onTaskDone` listener is still effect-scoped to `tool-bash`'s `apply`, so a completion landing during the reload gap still drops its one notice — the pre-existing reload-gap drop — but the ownership fence itself is HMR-proof.)
|
||||
Background-job ownership moved from a `tool-bash` plugin-local `Map<string, Agent>` into the executor. `ShellExecRequest` gains an optional `owner?: string`; the resolved `ShellExecSpec` carries it as required-but-nullable `owner: string | undefined` (a forgotten owner is a visible `undefined`, never a silently-absent property). The executor stores the token on its task and exposes it via a new `ShellExecutor.ownerOf(id): string | undefined` method (NOT on the public `BashTask` — one read path, no redundant API). `tool-bash` deletes its `Map` entirely: it stamps `exec.agent?.id` (the shared registry/session id) as the owner at `start`, and `bash_output`/`bash_kill` compare `ctx.shell.ownerOf(id)` to the caller's token with `!== undefined` semantics (an empty-string token is still a real owner). The completion notice finds the live agent by scanning `ctx.get('agents')?.list()` for `agent.id === ownerToken` (read via `ctx.get` — `onJobDone` runs on the bash fiber, a foreign fiber, where the `ctx.agents` proxy would throw). Because ownership now lives on the task in the executor (disposed with the `dsh-shell` fiber), it SURVIVES a `tool-bash` HMR reload — closing the old `XXX(tool-bash-owner-hmr)` gap. (The `onJobDone` listener is still effect-scoped to `tool-bash`'s `apply`, so a completion landing during the reload gap still drops its one notice — the pre-existing reload-gap drop — but the ownership fence itself is HMR-proof.)
|
||||
|
||||
## Verification
|
||||
|
||||
@@ -32,7 +32,7 @@ These invariants hold and are pinned by tests:
|
||||
|
||||
- ACP disconnect or plugin teardown leaves no registered agent and no session-store entry for any bridge-owned session, including a create racing connection closure.
|
||||
- `session/cancel` before a queued prompt starts prevents that prompt from running; a later accepted prompt remains an independent queued turn.
|
||||
- A `tool-bash` HMR reload does NOT make an existing background task readable or killable by a different session (ownership survives on the executor).
|
||||
- A `tool-bash` HMR reload does NOT make an existing background job readable or killable by a different session (ownership survives on the executor).
|
||||
- Existing non-ACP demos still work without managing handles explicitly; config-created agents remain owned by the `AgentLoop` plugin fiber.
|
||||
|
||||
## Session owner tokens are unique among live agents
|
||||
@@ -41,7 +41,7 @@ The bash owner-token comparison relies on the shared `Agent.id`/`SessionId` bein
|
||||
|
||||
## Alternatives considered
|
||||
|
||||
- **A public `BashTask.owner` field** instead of the `BashExecutor.ownerOf(id)` Service Definition method — rejected: one read path, no redundant API.
|
||||
- **A public `BashTask.owner` field** instead of the `ShellExecutor.ownerOf(id)` Service Definition method — rejected: one read path, no redundant API.
|
||||
- **Sibling cordis effects for the agent's session lifecycle** — rejected: a fiber unload disposes sibling effects concurrently (`Promise.all`), racing removal of the store-owned append publication hooks against the loop's closing `session/flush`; the single composite effect's ordered LIFO chain is what captures the closing `turn/end` on both disposal paths.
|
||||
- **A separate step-only `abort()` beside `cancel()`** — shipped originally, then removed as unused; `cancel()` is the single public stop primitive ([the public-stop-API Agent Note](../simplification/2026-06-20-public-agent-stop-api.md)).
|
||||
|
||||
|
||||
@@ -24,7 +24,7 @@ ACP（Agent Client Protocol）与 tool-bash 的若干限制是同一个所有权
|
||||
|
||||
### 3. Service Definition 中的 Bash 所有者令牌
|
||||
|
||||
后台任务所有权从 `tool-bash` 插件本地的 `Map<string, Agent>` 移入执行器。`BashExecRequest` 新增可选的 `owner?: string`；解析后的 `BashExecSpec` 将其作为必需但可空的 `owner: string | undefined` 携带（被遗忘的 owner 是可见的 `undefined`，而非静默缺失的属性）。执行器把 token 存在任务上，并通过新的 `BashExecutor.ownerOf(id): string | undefined` 方法暴露它（不放在公开的 `BashTask` 上——只有一条读取路径，没有冗余 API）。`tool-bash` 完全删除其 `Map`：它在 `start` 时将 `exec.agent?.id`（共享的注册表/会话 id）盖章为 owner，`bash_output`/`bash_kill` 则以 `!== undefined` 语义把 `ctx.bash.ownerOf(id)` 与调用方 token 比较（空字符串 token 仍是真实 owner）。完成通知通过扫描 `ctx.get('agents')?.list()` 查找 `agent.id === ownerToken` 的存活 agent（经 `ctx.get` 读取——`onTaskDone` 运行在 bash fiber 这一外部 fiber 上，直接使用 `ctx.agents` proxy 会抛异常）。由于所有权现在保存在执行器的任务上（随 `dsh-bash` fiber dispose），它能跨越 `tool-bash` HMR 重载，关闭旧的 `XXX(tool-bash-owner-hmr)` 缺口。（`onTaskDone` 监听器仍受 `tool-bash` 的 `apply` effect 约束，因此落在重载间隙的完成仍会丢失一条通知——既有的重载间隙丢失——但所有权隔离本身已经不受 HMR 影响。）
|
||||
后台任务所有权从 `tool-bash` 插件本地的 `Map<string, Agent>` 移入执行器。`ShellExecRequest` 新增可选的 `owner?: string`；解析后的 `ShellExecSpec` 将其作为必需但可空的 `owner: string | undefined` 携带（被遗忘的 owner 是可见的 `undefined`，而非静默缺失的属性）。执行器把 token 存在任务上，并通过新的 `ShellExecutor.ownerOf(id): string | undefined` 方法暴露它（不放在公开的 `BashTask` 上——只有一条读取路径，没有冗余 API）。`tool-bash` 完全删除其 `Map`：它在 `start` 时将 `exec.agent?.id`（共享的注册表/会话 id）盖章为 owner，`bash_output`/`bash_kill` 则以 `!== undefined` 语义把 `ctx.shell.ownerOf(id)` 与调用方 token 比较（空字符串 token 仍是真实 owner）。完成通知通过扫描 `ctx.get('agents')?.list()` 查找 `agent.id === ownerToken` 的存活 agent（经 `ctx.get` 读取——`onJobDone` 运行在 bash fiber 这一外部 fiber 上，直接使用 `ctx.agents` proxy 会抛异常）。由于所有权现在保存在执行器的任务上（随 `dsh-shell` fiber dispose），它能跨越 `tool-bash` HMR 重载，关闭旧的 `XXX(tool-bash-owner-hmr)` 缺口。（`onJobDone` 监听器仍受 `tool-bash` 的 `apply` effect 约束，因此落在重载间隙的完成仍会丢失一条通知——既有的重载间隙丢失——但所有权隔离本身已经不受 HMR 影响。）
|
||||
|
||||
## 验证
|
||||
|
||||
@@ -41,7 +41,7 @@ bash 所有者 token 比较依赖共享的 `Agent.id`/`SessionId` 在存活 agen
|
||||
|
||||
## 曾考虑的替代方案
|
||||
|
||||
- **公开的 `BashTask.owner` 字段**而非 `BashExecutor.ownerOf(id)` Service Definition 方法：否决。一条读取路径即可，无需冗余 API。
|
||||
- **公开的 `BashTask.owner` 字段**而非 `ShellExecutor.ownerOf(id)` Service Definition 方法：否决。一条读取路径即可，无需冗余 API。
|
||||
- **为 agent 的会话生命周期使用兄弟 Cordis effect**：否决。fiber 卸载时并发释放兄弟 effect（`Promise.all`），store 拥有的 append 发布钩子的移除与循环的关闭 `session/flush` 产生竞争；单一复合 effect 的有序 LIFO 链才能在两条释放路径上都捕获关闭的 `turn/end`。
|
||||
- **在 `cancel()` 之外另设一个仅中止步骤的 `abort()`**：最初发布过，后因无人使用而移除；`cancel()` 是唯一的公开停止原语（见[公开停止接口 Agent Note](../simplification/2026-06-20-public-agent-stop-api.md)）。
|
||||
|
||||
|
||||
@@ -2,5 +2,5 @@
|
||||
# side as of the last confirmed-consistent state. Both languages carry equal authority;
|
||||
# after editing either side, bring the other along and re-record with:
|
||||
# pnpm run verify-translation-pairing --write .agents/notes/implemented/architecture/2026-06-18-session-surface.md
|
||||
2026-06-18-session-surface.md: fd48a4ee4e177b84a235b044b054e4ac266df8ba
|
||||
2026-06-18-session-surface.zh.md: 496624016cd68e2e8b0fe65ed0246c5d469c4585
|
||||
2026-06-18-session-surface.md: 3682ae7b8b58b9e5d40695732c3a1531d0651d5e
|
||||
2026-06-18-session-surface.zh.md: 8cba9645dc6d0c8a4d1ee096668fc0bc38aaa725
|
||||
|
||||
@@ -68,6 +68,6 @@ Every surface-eligible event must carry `surfaceOp` or it would disappear from d
|
||||
- **`packages/session/session-persistence-jsonl`**: No changes required.
|
||||
- **`packages/session/session-persistence`**: Abstract interface unchanged.
|
||||
|
||||
The surface is the foundation history manipulation ships on — dsh-compact's compaction rides it. A compaction or tool-result-prune plugin appends one of the existing message-producing event types (a `user/message` carrying the summary, say) with `surfaceOp: { op: 'replace', start, end }` and `sourceEventSeqs` covering the shadowed entries — the new event takes the range's place on the surface while the plugin's own trace events (e.g. `compaction/start`, `compaction/end`) stay off it. Replay preserves the decision deterministically.
|
||||
The surface is the foundation history manipulation ships on — dsh-compaction's compaction rides it. A compaction or tool-result-pruner plugin appends one of the existing message-producing event types (a `user/message` carrying the summary, say) with `surfaceOp: { op: 'replace', start, end }` and `sourceEventSeqs` covering the shadowed entries — the new event takes the range's place on the surface while the plugin's own trace events (e.g. `compaction/start`, `compaction/end`) stay off it. Replay preserves the decision deterministically.
|
||||
|
||||
A `tool/result` replacement may rewrite exactly one current `tool/result` and must preserve every data field except `content`. Session acceptance enforces this rule together with positional range and cited source-event validation, independent of optional diagnostic plugins.
|
||||
|
||||
@@ -68,6 +68,6 @@ export type SurfaceOp =
|
||||
- **`packages/session/session-persistence-jsonl`**：无需改动。
|
||||
- **`packages/session/session-persistence`**：抽象接口不变。
|
||||
|
||||
surface 是历史操纵赖以落地的基础——dsh-compact 的压缩就搭载于其上。压缩或 tool-result-prune 插件追加一个既有的消息产出事件类型（例如一条携带摘要的 `user/message`），附带 `surfaceOp: { op: 'replace', start, end }` 和覆盖被遮蔽条目的 `sourceEventSeqs`——新事件在 surface 上取代该范围的位置，而插件自身的 trace 事件（如 `compaction/start`、`compaction/end`）不进入 surface。回放以确定性方式保留该决策。
|
||||
surface 是历史操纵赖以落地的基础——dsh-compaction 的压缩就搭载于其上。压缩或 tool-result-pruner 插件追加一个既有的消息产出事件类型（例如一条携带摘要的 `user/message`），附带 `surfaceOp: { op: 'replace', start, end }` 和覆盖被遮蔽条目的 `sourceEventSeqs`——新事件在 surface 上取代该范围的位置，而插件自身的 trace 事件（如 `compaction/start`、`compaction/end`）不进入 surface。回放以确定性方式保留该决策。
|
||||
|
||||
一次 `tool/result` 替换只能改写当前的一个 `tool/result`，并且必须保留除 `content` 以外的每个数据字段。Session 接纳会与位置范围和引用的源事件校验一起强制这条规则，不依赖可选的诊断插件。
|
||||
|
||||
@@ -2,5 +2,5 @@
|
||||
# side as of the last confirmed-consistent state. Both languages carry equal authority;
|
||||
# after editing either side, bring the other along and re-record with:
|
||||
# pnpm run verify-translation-pairing --write .agents/notes/implemented/architecture/2026-06-20-branded-ids.md
|
||||
2026-06-20-branded-ids.md: ded48409bcb3deb19e35029fa795b5ea28c9f6d7
|
||||
2026-06-20-branded-ids.zh.md: 824d802b4aaec9f4a6ae530659387126a30fbc2d
|
||||
2026-06-20-branded-ids.md: 6bf68c9b6b6dea8317992bcb6f694a8ee141f048
|
||||
2026-06-20-branded-ids.zh.md: 11a78df609888fdf917246940c12463373ff2de3
|
||||
|
||||
@@ -8,9 +8,9 @@ English | [中文](2026-06-20-branded-ids.zh.md)
|
||||
|
||||
The harness brands `CallId` (`packages/llm/llm/src/brand.ts`) and the shared agent/session `SessionId` (`packages/core/session/src/types.ts`) using the `Branded<B> = string & { readonly [BRAND]: B }` machinery (owned by the type-only `@deepseek-ai/dsh-brand` package at `packages/util/brand/` — see its [README](../../../../packages/util/brand/README.md)) and a zero-cost cast factory per type. `dsh-brand` also states the governing policy: *"Branding is for ids that cross package boundaries and could plausibly be confused; not every string needs a brand."* That policy is right; the problem is that it is only half-applied. Two gaps let a structurally-identical-but-semantically-wrong string slip through the type checker today.
|
||||
|
||||
**Gap 1 — unbranded cross-boundary IDs in the bash seam.** The background-task id is a plain `string`: `BashTask.id: string` (`packages/bash/bash/src/types.ts`), carried as `string` through the whole executor seam (`BashExecutor.get`/`ownerOf`/`readOutput`/`kill(id: string)` in `packages/bash/bash/src/index.ts`) and validated/passed as `string` by the model-facing tools (`validateTaskId`, `assertTaskAccess`, the `task_id` schema arg in `packages/bash/tool-bash/src/index.ts`). It is generated by a per-executor counter — `` `bash-${this.nextTaskId++}` `` in `packages/bash/bash-local/src/index.ts` — which gives it **exactly the same `name-N` shape as `SessionId`'s default** (`` `session-${++counter}` `` in `packages/core/session/src/index.ts`). A bash task id and a session id are trivially swappable at a call site and the compiler says nothing. It is a model-facing id (the model passes `task_id` back to `bash_output`/`bash_kill`), so a confusion here is reachable from untrusted input.
|
||||
**Gap 1 — unbranded cross-boundary IDs in the bash seam.** The background-job id is a plain `string`: `BashTask.id: string` (`packages/shell/shell/src/types.ts`), carried as `string` through the whole executor seam (`ShellExecutor.get`/`ownerOf`/`readOutput`/`kill(id: string)` in `packages/shell/shell/src/index.ts`) and validated/passed as `string` by the model-facing tools (`validateJobId`, `assertTaskAccess`, the `job_id` schema arg in `packages/shell/tool-bash/src/index.ts`). It is generated by a per-executor counter — `` `bash-${this.nextTaskId++}` `` in `packages/shell/bash-local/src/index.ts` — which gives it **exactly the same `name-N` shape as `SessionId`'s default** (`` `session-${++counter}` `` in `packages/core/session/src/index.ts`). A bash job id and a session id are trivially swappable at a call site and the compiler says nothing. It is a model-facing id (the model passes `job_id` back to `bash_output`/`bash_kill`), so a confusion here is reachable from untrusted input.
|
||||
|
||||
The bash **owner token** is the related sub-case: `BashExecRequest.owner?: string` and `BashExecSpec.owner: string | undefined` (`packages/bash/bash/src/types.ts`) are documented as a deliberately *opaque* isolation key, but in every live caller the value IS the owning agent's shared `Agent.id`/`SessionId` (`callerToken = (exec) => exec.agent?.id` in `packages/bash/tool-bash/src/index.ts`) wearing a different seam-local name. It is compared for access control (`owner !== callerToken(exec)`), so a mismatched-but-well-typed string here is a cross-session isolation bug the type system currently cannot catch. This is the shared id alias covered by the [unified agent/session identity decision](../simplification/2026-06-20-unify-agent-and-session-id.md).
|
||||
The bash **owner token** is the related sub-case: `ShellExecRequest.owner?: string` and `ShellExecSpec.owner: string | undefined` (`packages/shell/shell/src/types.ts`) are documented as a deliberately *opaque* isolation key, but in every live caller the value IS the owning agent's shared `Agent.id`/`SessionId` (`callerToken = (exec) => exec.agent?.id` in `packages/shell/tool-bash/src/index.ts`) wearing a different seam-local name. It is compared for access control (`owner !== callerToken(exec)`), so a mismatched-but-well-typed string here is a cross-session isolation bug the type system currently cannot catch. This is the shared id alias covered by the [unified agent/session identity decision](../simplification/2026-06-20-unify-agent-and-session-id.md).
|
||||
|
||||
**Gap 2 — brand erosion at the boundaries of the *already-branded* IDs.** Even `CallId` and `SessionId` decay back to bare `string` at exactly the places confusion is most likely: registry/store key types and public method params. Representative sites include the session store, the agent registry (both keyed by the shared `SessionId`), tool-presentation call-id maps, ACP's session records, and the persistence coordinator. A brand that is dropped at a collection key buys nothing on lookups — the value of the existing brands is partly unrealized.
|
||||
|
||||
@@ -18,9 +18,9 @@ The bash **owner token** is the related sub-case: `BashExecRequest.owner?: strin
|
||||
|
||||
A type-only change. Brands are zero-cost casts; nothing about runtime behavior, serialization, comparison, or the wire format changes. The decision has three parts, all honoring the existing "not every string" policy.
|
||||
|
||||
- **Brand the bash task id.** Add `BashTaskId = Branded<'BashTaskId'>` plus its same-named factory in `packages/bash/bash/src/types.ts` (the package that *owns* the id), importing `Branded` from `@deepseek-ai/dsh-brand` exactly as `SessionId` does. The brand primitive lives in the dependency-free `dsh-brand` utility package precisely so `dsh-bash` can brand its ids by depending on it alone — it never pulls in `dsh-llm` (or `dsh-session`) just to reach `Branded`. Thread it through `BashTask.id`, the `BashExecutor` Service Definition methods (`get`/`ownerOf`/`readOutput`/`kill`), the generation site in `dsh-bash-local` (brand the counter output once, at creation), and the `dsh-tool-bash` validate/access surface (`validateTaskId` returns a `BashTaskId`; `task_id` is branded at the tool boundary where the model's string arrives).
|
||||
- **Brand the bash job id.** Add `BashTaskId = Branded<'BashTaskId'>` plus its same-named factory in `packages/shell/shell/src/types.ts` (the package that *owns* the id), importing `Branded` from `@deepseek-ai/dsh-brand` exactly as `SessionId` does. The brand primitive lives in the dependency-free `dsh-brand` utility package precisely so `dsh-shell` can brand its ids by depending on it alone — it never pulls in `dsh-llm` (or `dsh-session`) just to reach `Branded`. Thread it through `BashTask.id`, the `ShellExecutor` Service Definition methods (`get`/`ownerOf`/`readOutput`/`kill`), the generation site in `dsh-bash-local` (brand the counter output once, at creation), and the `dsh-tool-bash` validate/access surface (`validateJobId` returns a `BashTaskId`; `job_id` is branded at the tool boundary where the model's string arrives).
|
||||
|
||||
- **Mint a distinct `OwnerToken` brand.** Add `OwnerToken = Branded<'OwnerToken'>` in `packages/bash/bash/src/types.ts`; type `BashExecRequest.owner` / `BashExecSpec.owner` / `BashExecutor.ownerOf` as `OwnerToken | undefined`. The `dsh-tool-bash` consumer casts the agent's shared `id` (`SessionId`) into an `OwnerToken` at the boundary — the one place the two vocabularies meet. The bash Service Definition never imports `dsh-session`. (Rationale in the next section.)
|
||||
- **Mint a distinct `OwnerToken` brand.** Add `OwnerToken = Branded<'OwnerToken'>` in `packages/shell/shell/src/types.ts`; type `ShellExecRequest.owner` / `ShellExecSpec.owner` / `ShellExecutor.ownerOf` as `OwnerToken | undefined`. The `dsh-tool-bash` consumer casts the agent's shared `id` (`SessionId`) into an `OwnerToken` at the boundary — the one place the two vocabularies meet. The bash Service Definition never imports `dsh-session`. (Rationale in the next section.)
|
||||
|
||||
- **Stop the brand erosion.** Propagate the existing brands to the `Map` key types and public method params listed under Gap 2 — `Map<SessionId, Session>`, `Map<SessionId, Agent>`, `get(id: SessionId)`, `Map<CallId, …>`, ACP's `SessionId` surface, and the coordinator's `Map<SessionId, …>`. This is the larger mechanical share of the change and the part that makes the *existing* brands actually load-bearing on lookups, not just on struct fields.
|
||||
|
||||
@@ -46,21 +46,21 @@ export function OwnerToken(id: string): OwnerToken {
|
||||
|
||||
### Why not typing `owner` as `SessionId`?
|
||||
|
||||
The obvious shortcut is to type `owner` as `SessionId` directly — it always *is* one. We reject that. The bash executor seam is a capability seam (Service Definition `dsh-bash`, Service provider `dsh-bash-local`, Consumer `dsh-tool-bash`) and its owner token is *documented as deliberately opaque*: the executor "never interprets it (no access policy lives in the seam — that is the consumer's job)" (`packages/bash/bash/src/types.ts`). Typing the Service Definition's field as `SessionId` would import `dsh-session`'s vocabulary into a package that must not know what an owner token *means* — it would couple a generic execution backend to the session model and contradict the opaque-token design. A sandboxed or remote executor that replaces `dsh-bash-local` should not inherit a session dependency. The distinct `OwnerToken` brand keeps the seam decoupled: `dsh-bash` knows only "an owner is some opaque branded token," and the `dsh-tool-bash` consumer — which already decides the access policy — is the single boundary that casts its `SessionId` into an `OwnerToken`. The brand still delivers the safety win (you cannot pass a `BashTaskId` or a raw string where an owner is expected) without the coupling.
|
||||
The obvious shortcut is to type `owner` as `SessionId` directly — it always *is* one. We reject that. The bash executor seam is a capability seam (Service Definition `dsh-shell`, Service provider `dsh-bash-local`, Consumer `dsh-tool-bash`) and its owner token is *documented as deliberately opaque*: the executor "never interprets it (no access policy lives in the seam — that is the consumer's job)" (`packages/shell/shell/src/types.ts`). Typing the Service Definition's field as `SessionId` would import `dsh-session`'s vocabulary into a package that must not know what an owner token *means* — it would couple a generic execution backend to the session model and contradict the opaque-token design. A sandboxed or remote executor that replaces `dsh-bash-local` should not inherit a session dependency. The distinct `OwnerToken` brand keeps the seam decoupled: `dsh-shell` knows only "an owner is some opaque branded token," and the `dsh-tool-bash` consumer — which already decides the access policy — is the single boundary that casts its `SessionId` into an `OwnerToken`. The brand still delivers the safety win (you cannot pass a `BashTaskId` or a raw string where an owner is expected) without the coupling.
|
||||
|
||||
## Out of scope / possible extensions
|
||||
|
||||
Kept deliberately narrow per the "not every string needs a brand" policy. Each of these is a plausible future brand, deferred with a reason, not a commitment:
|
||||
|
||||
- **`ModelId`** (`GenerateOptions.model`, the `LlmService` adapter-registry key) — a real cross-package lookup key (config → agent → llm → adapter); a reasonable next brand, left out only to keep this decision's blast radius focused.
|
||||
- **`ToolName`** (the `ToolRegistry` key) — author-defined, human-readable, and rarely confused with another id; the weakest candidate, likely not worth a brand.
|
||||
- **`ModelId`** (`GenerateOptions.model`, the `LlmRuntime` adapter-registry key) — a real cross-package lookup key (config → agent → llm → adapter); a reasonable next brand, left out only to keep this decision's blast radius focused.
|
||||
- **`ToolName`** (the `ToolRuntime` key) — author-defined, human-readable, and rarely confused with another id; the weakest candidate, likely not worth a brand.
|
||||
- **`ErrorCode`** (`HarnessError.code`) — a closed vocabulary (`ABORTED`, `NO_ADAPTER`, …), not a per-instance id; better served by a string-literal union than a brand, if anything.
|
||||
- **Numeric ordinals** — turn number, step number, and the event `seq` are `number`, not `string`, so `Branded<string>` does not apply; a parallel `number & { readonly [BRAND]: B }` variant could brand them, but they are positional ordinals rarely passed across boundaries, so the payoff is low.
|
||||
- **Validated construction** — the brand factories are pure casts with no runtime check, and every boundary (ACP `sessionId`, provider-issued `call.id`, the empty-string fallback in `dsh-llm-deepseek`) trusts the raw string today. A `SessionId.parse()` / `isValid()` companion that throws on malformed input at boundaries is a genuine gap, but it is a *runtime-behavior* change with its own design (what is "malformed"? what happens on failure?) and belongs in its own decision, not bundled into this type-only change.
|
||||
|
||||
## Verification
|
||||
|
||||
The landed invariants: `BashTaskId` and `OwnerToken` are defined in `dsh-bash` and threaded end-to-end (Service Definition, the `dsh-bash-local` generation site, the `dsh-tool-bash` model-facing tool) with no `dsh-bash` dependency on `dsh-session`; no collection keyed by an in-scope branded id (`CallId`/`SessionId`/`BashTaskId`) is keyed by bare `string`; public method params and exported signatures keep the brand; and brands are constructed via the cast factory at each boundary where a raw string enters (provider call id, ACP session id, model-supplied `task_id`), never as scattered `as` casts.
|
||||
The landed invariants: `BashTaskId` and `OwnerToken` are defined in `dsh-shell` and threaded end-to-end (Service Definition, the `dsh-bash-local` generation site, the `dsh-tool-bash` model-facing tool) with no `dsh-shell` dependency on `dsh-session`; no collection keyed by an in-scope branded id (`CallId`/`SessionId`/`BashTaskId`) is keyed by bare `string`; public method params and exported signatures keep the brand; and brands are constructed via the cast factory at each boundary where a raw string enters (provider call id, ACP session id, model-supplied `job_id`), never as scattered `as` casts.
|
||||
|
||||
## Consequences
|
||||
|
||||
|
||||
@@ -8,9 +8,9 @@ Status: implemented
|
||||
|
||||
harness 使用 `Branded<B> = string & { readonly [BRAND]: B }` 机制，为 `CallId`（`packages/llm/llm/src/brand.ts`）和 agent（智能体）/会话共享的 `SessionId`（`packages/core/session/src/types.ts`）做 brand 处理；该机制由纯类型包 `@deepseek-ai/dsh-brand` 拥有，位于 `packages/util/brand/`，见其 [README](../../../../packages/util/brand/README.md)，并为每个类型提供零开销的 cast 工厂。`dsh-brand` 还声明了治理策略：*「Branding 用于跨包边界且可能被混淆的 id；不是每个 string 都需要 brand。」* 这条策略是正确的；问题在于它只落实了一半。两处缺口使得结构相同但语义错误的 string 今天仍能通过类型检查器。
|
||||
|
||||
**缺口 1：bash seam 中未 brand 的跨边界 ID。** 后台 task id 是普通 `string`：`BashTask.id: string`（`packages/bash/bash/src/types.ts`），作为 `string` 贯穿整个执行器 seam（`packages/bash/bash/src/index.ts` 中的 `BashExecutor.get`/`ownerOf`/`readOutput`/`kill(id: string)`），再由面向模型的工具以 `string` 校验并传递（`validateTaskId`、`assertTaskAccess`、`packages/bash/tool-bash/src/index.ts` 中 `task_id` 的 schema 参数）。它由每执行器计数器生成——`packages/bash/bash-local/src/index.ts` 中的 `` `bash-${this.nextTaskId++}` ``——其形状与 `SessionId` 的默认值**完全相同，都是 `name-N`**（`packages/core/session/src/index.ts` 中的 `` `session-${++counter}` ``）。bash task id 和会话 id 在调用点轻易就能互换，而编译器毫无反应。它是面向模型的 id（模型会把 `task_id` 传回 `bash_output`/`bash_kill`），所以该混淆可由不受信任的输入触达。
|
||||
**缺口 1：bash seam 中未 brand 的跨边界 ID。** 后台 job id 是普通 `string`：`BashTask.id: string`（`packages/shell/shell/src/types.ts`），作为 `string` 贯穿整个执行器 seam（`packages/shell/shell/src/index.ts` 中的 `ShellExecutor.get`/`ownerOf`/`readOutput`/`kill(id: string)`），再由面向模型的工具以 `string` 校验并传递（`validateJobId`、`assertTaskAccess`、`packages/shell/tool-bash/src/index.ts` 中 `job_id` 的 schema 参数）。它由每执行器计数器生成——`packages/shell/bash-local/src/index.ts` 中的 `` `bash-${this.nextTaskId++}` ``——其形状与 `SessionId` 的默认值**完全相同，都是 `name-N`**（`packages/core/session/src/index.ts` 中的 `` `session-${++counter}` ``）。bash job id 和会话 id 在调用点轻易就能互换，而编译器毫无反应。它是面向模型的 id（模型会把 `job_id` 传回 `bash_output`/`bash_kill`），所以该混淆可由不受信任的输入触达。
|
||||
|
||||
bash **owner token** 是相关的子情形：`BashExecRequest.owner?: string` 和 `BashExecSpec.owner: string | undefined`（`packages/bash/bash/src/types.ts`）被文档描述为刻意*不透明*的隔离键，但在所有实际调用方中，该值就是所属 agent 共享的 `Agent.id`/`SessionId`（`callerToken = (exec) => exec.agent?.id`，位于 `packages/bash/tool-bash/src/index.ts`），只是披着另一个 seam 本地名称。它被用于访问控制比较（`owner !== callerToken(exec)`），因此一个不匹配但类型正确的 string 在此处就是跨会话隔离 bug，而当前类型系统无法捕获。这正是[统一 agent/session 标识决策](../simplification/2026-06-20-unify-agent-and-session-id.md)覆盖的共享 id 别名。
|
||||
bash **owner token** 是相关的子情形：`ShellExecRequest.owner?: string` 和 `ShellExecSpec.owner: string | undefined`（`packages/shell/shell/src/types.ts`）被文档描述为刻意*不透明*的隔离键，但在所有实际调用方中，该值就是所属 agent 共享的 `Agent.id`/`SessionId`（`callerToken = (exec) => exec.agent?.id`，位于 `packages/shell/tool-bash/src/index.ts`），只是披着另一个 seam 本地名称。它被用于访问控制比较（`owner !== callerToken(exec)`），因此一个不匹配但类型正确的 string 在此处就是跨会话隔离 bug，而当前类型系统无法捕获。这正是[统一 agent/session 标识决策](../simplification/2026-06-20-unify-agent-and-session-id.md)覆盖的共享 id 别名。
|
||||
|
||||
**缺口 2：*已经 brand* 的 ID 在边界处被侵蚀。** 就连 `CallId` 和 `SessionId` 也恰好在最容易混淆的地方退化为裸 `string`：注册表/store 键类型和公开方法参数。代表性位置包括会话存储、agent 注册表（二者都以共享的 `SessionId` 为键）、工具展示层的 call-id map、ACP（Agent Client Protocol）的会话记录，以及持久化协调器。在集合键处丢弃 brand，会让既有 brand 在查找时毫无价值；它们的价值只实现了一部分。
|
||||
|
||||
@@ -18,9 +18,9 @@ bash **owner token** 是相关的子情形：`BashExecRequest.owner?: string`
|
||||
|
||||
纯类型变更。Brand 是零开销 cast；运行时行为、序列化、比较和协议格式（wire format）均不变。该决策分三部分，全部遵循既有的「不是每个 string 都需要」策略。
|
||||
|
||||
- **为 bash task id 加 brand。** 在 `packages/bash/bash/src/types.ts`（*拥有*该 id 的包）中添加 `BashTaskId = Branded<'BashTaskId'>` 及其同名工厂，从 `@deepseek-ai/dsh-brand` 导入 `Branded`，方式与 `SessionId` 完全一致。brand 原语位于无依赖的 `dsh-brand` 工具包中，正是为了让 `dsh-bash` 仅依赖它就能为自己的 id 加 brand，而无需引入 `dsh-llm`（或 `dsh-session`）来获取 `Branded`。将其贯穿 `BashTask.id`、`BashExecutor` Service Definition 方法（`get`/`ownerOf`/`readOutput`/`kill`）、`dsh-bash-local` 中的生成点（在创建时对计数器输出做一次 brand），以及 `dsh-tool-bash` 的校验/访问面（`validateTaskId` 返回 `BashTaskId`；`task_id` 在模型 string 到达的工具边界处被 brand）。
|
||||
- **为 bash job id 加 brand。** 在 `packages/shell/shell/src/types.ts`（*拥有*该 id 的包）中添加 `BashTaskId = Branded<'BashTaskId'>` 及其同名工厂，从 `@deepseek-ai/dsh-brand` 导入 `Branded`，方式与 `SessionId` 完全一致。brand 原语位于无依赖的 `dsh-brand` 工具包中，正是为了让 `dsh-shell` 仅依赖它就能为自己的 id 加 brand，而无需引入 `dsh-llm`（或 `dsh-session`）来获取 `Branded`。将其贯穿 `BashTask.id`、`ShellExecutor` Service Definition 方法（`get`/`ownerOf`/`readOutput`/`kill`）、`dsh-bash-local` 中的生成点（在创建时对计数器输出做一次 brand），以及 `dsh-tool-bash` 的校验/访问面（`validateJobId` 返回 `BashTaskId`；`job_id` 在模型 string 到达的工具边界处被 brand）。
|
||||
|
||||
- **铸造独立的 `OwnerToken` brand。** 在 `packages/bash/bash/src/types.ts` 中添加 `OwnerToken = Branded<'OwnerToken'>`；将 `BashExecRequest.owner` / `BashExecSpec.owner` / `BashExecutor.ownerOf` 的类型标注为 `OwnerToken | undefined`。`dsh-tool-bash` 消费方在边界处将 agent 共享的 `id`（`SessionId`）cast 为 `OwnerToken`——这是两套词汇唯一交汇的地方。bash Service Definition 从不导入 `dsh-session`。（理由见下一节。）
|
||||
- **铸造独立的 `OwnerToken` brand。** 在 `packages/shell/shell/src/types.ts` 中添加 `OwnerToken = Branded<'OwnerToken'>`；将 `ShellExecRequest.owner` / `ShellExecSpec.owner` / `ShellExecutor.ownerOf` 的类型标注为 `OwnerToken | undefined`。`dsh-tool-bash` 消费方在边界处将 agent 共享的 `id`（`SessionId`）cast 为 `OwnerToken`——这是两套词汇唯一交汇的地方。bash Service Definition 从不导入 `dsh-session`。（理由见下一节。）
|
||||
|
||||
- **阻止 brand 侵蚀。** 将既有 brand 传播到缺口 2 列出的 `Map` 键类型和公开方法参数中：`Map<SessionId, Session>`、`Map<SessionId, Agent>`、`get(id: SessionId)`、`Map<CallId, …>`、ACP 的 `SessionId` surface、协调器的 `Map<SessionId, …>`。这是变更中机械量最大的部分，也是让*既有* brand 在查找处真正发挥作用（而不仅仅标注在结构体字段上）的关键。
|
||||
|
||||
@@ -46,21 +46,21 @@ export function OwnerToken(id: string): OwnerToken {
|
||||
|
||||
### 为什么不把 `owner` 类型标注为 `SessionId`？
|
||||
|
||||
显而易见的捷径是直接把 `owner` 类型标注为 `SessionId`——它确实*总是*一个会话 id。我们否决这个方案。bash 执行器 seam 是能力 seam（Service Definition `dsh-bash`、Service provider `dsh-bash-local`、Consumer `dsh-tool-bash`），其 owner token 被*明确记录为刻意不透明*：执行器「从不解释它（seam 中没有访问策略——那是消费方的职责）」（`packages/bash/bash/src/types.ts`）。把 Service Definition 的字段类型标注为 `SessionId`，会把 `dsh-session` 的词汇引入一个不应知道 owner token *含义*的包——这会让通用执行后端耦合会话模型，并违背不透明 token 的设计。取代 `dsh-bash-local` 的沙箱化执行器或远程执行器不应继承会话依赖。独立的 `OwnerToken` brand 使 seam 保持解耦：`dsh-bash` 只知道「owner 是某种带 brand 的不透明 token」，而已经决定访问策略的 `dsh-tool-bash` 消费方，是把其 `SessionId` cast 为 `OwnerToken` 的唯一边界。该 brand 仍带来安全收益（不能把 `BashTaskId` 或裸 string 传到 owner 位置），且不引入耦合。
|
||||
显而易见的捷径是直接把 `owner` 类型标注为 `SessionId`——它确实*总是*一个会话 id。我们否决这个方案。bash 执行器 seam 是能力 seam（Service Definition `dsh-shell`、Service provider `dsh-bash-local`、Consumer `dsh-tool-bash`），其 owner token 被*明确记录为刻意不透明*：执行器「从不解释它（seam 中没有访问策略——那是消费方的职责）」（`packages/shell/shell/src/types.ts`）。把 Service Definition 的字段类型标注为 `SessionId`，会把 `dsh-session` 的词汇引入一个不应知道 owner token *含义*的包——这会让通用执行后端耦合会话模型，并违背不透明 token 的设计。取代 `dsh-bash-local` 的沙箱化执行器或远程执行器不应继承会话依赖。独立的 `OwnerToken` brand 使 seam 保持解耦：`dsh-shell` 只知道「owner 是某种带 brand 的不透明 token」，而已经决定访问策略的 `dsh-tool-bash` 消费方，是把其 `SessionId` cast 为 `OwnerToken` 的唯一边界。该 brand 仍带来安全收益（不能把 `BashTaskId` 或裸 string 传到 owner 位置），且不引入耦合。
|
||||
|
||||
## 不在范围内 / 可能的扩展
|
||||
|
||||
遵循「不是每个 string 都需要 brand」的策略，刻意保持窄范围。以下每项都是合理的未来 brand 候选，附带推迟理由而非承诺：
|
||||
|
||||
- **`ModelId`**（`GenerateOptions.model`，`LlmService` 适配器注册表的键）：一个真正的跨包查找键（config → agent → llm → 适配器）；合理的下一个 brand，仅为控制本决策的影响范围而暂不纳入。
|
||||
- **`ToolName`**（`ToolRegistry` 的键）：由作者定义、人类可读，且很少与其他 id 混淆；最弱的候选，可能不值得加 brand。
|
||||
- **`ModelId`**（`GenerateOptions.model`，`LlmRuntime` 适配器注册表的键）：一个真正的跨包查找键（config → agent → llm → 适配器）；合理的下一个 brand，仅为控制本决策的影响范围而暂不纳入。
|
||||
- **`ToolName`**（`ToolRuntime` 的键）：由作者定义、人类可读，且很少与其他 id 混淆；最弱的候选，可能不值得加 brand。
|
||||
- **`ErrorCode`**（`HarnessError.code`）：一个封闭词汇（`ABORTED`、`NO_ADAPTER`……），不是逐实例的 id；如果要做，string 字面量联合类型比 brand 更合适。
|
||||
- **数值序号**：轮次号、步骤号和事件 `seq` 是 `number` 而非 `string`，`Branded<string>` 不适用；可以用并行的 `number & { readonly [BRAND]: B }` 变体来 brand 它们，但它们是位置序号、很少跨边界传递，收益较低。
|
||||
- **带校验的构造**：brand 工厂是纯 cast，无运行时检查，且每个边界（ACP `sessionId`、提供方签发的 `call.id`、`dsh-llm-deepseek` 中的空字符串回退）今天都信任裸 string。一个在边界处对格式错误的输入抛异常的 `SessionId.parse()` / `isValid()` 配套工具确实是缺口，但它是*运行时行为*变更，有自己的设计问题（什么算「格式错误」？失败时会怎样？），应在独立决策中处理，不应捆绑进这次纯类型变更。
|
||||
|
||||
## 验证
|
||||
|
||||
已落地的不变式如下：`BashTaskId` 和 `OwnerToken` 定义在 `dsh-bash` 中，并端到端贯穿 Service Definition、`dsh-bash-local` 生成点与 `dsh-tool-bash` 面向模型的工具，且 `dsh-bash` 未添加对 `dsh-session` 的依赖；没有任何以范围内 brand id（`CallId`/`SessionId`/`BashTaskId`）为键的集合使用裸 `string`；公开方法参数和导出签名保留 brand；每个原始 string 进入的边界（提供方 call id、ACP 会话 id、模型提供的 `task_id`）都通过 cast 工厂构造 brand，而不是散落的 `as` cast。
|
||||
已落地的不变式如下：`BashTaskId` 和 `OwnerToken` 定义在 `dsh-shell` 中，并端到端贯穿 Service Definition、`dsh-bash-local` 生成点与 `dsh-tool-bash` 面向模型的工具，且 `dsh-shell` 未添加对 `dsh-session` 的依赖；没有任何以范围内 brand id（`CallId`/`SessionId`/`BashTaskId`）为键的集合使用裸 `string`；公开方法参数和导出签名保留 brand；每个原始 string 进入的边界（提供方 call id、ACP 会话 id、模型提供的 `job_id`）都通过 cast 工厂构造 brand，而不是散落的 `as` cast。
|
||||
|
||||
## 后果
|
||||
|
||||
|
||||
@@ -2,5 +2,5 @@
|
||||
# side as of the last confirmed-consistent state. Both languages carry equal authority;
|
||||
# after editing either side, bring the other along and re-record with:
|
||||
# pnpm run verify-translation-pairing --write .agents/notes/implemented/architecture/2026-06-20-generic-long-running-tool-runtime.md
|
||||
2026-06-20-generic-long-running-tool-runtime.md: 1edc3422c253e06178a5c8ebf68dfd4ef1289e31
|
||||
2026-06-20-generic-long-running-tool-runtime.zh.md: 1f5349b2aeb02e8db21150012300a6d0a0493ae1
|
||||
2026-06-20-generic-long-running-tool-runtime.md: 7db43323dd83f8e99698a7163412c9b33a607dfb
|
||||
2026-06-20-generic-long-running-tool-runtime.zh.md: fe12ef264d294224fcfaf503ba958fa90c782347
|
||||
|
||||
@@ -1,4 +1,4 @@
|
||||
# Agent Note: The background task runtime (`ctx.tasks`) and generic task control tools
|
||||
# Agent Note: The background job runtime (`ctx.jobs`) and generic task control tools
|
||||
|
||||
Status: implemented
|
||||
|
||||
@@ -6,38 +6,38 @@ English | [中文](2026-06-20-generic-long-running-tool-runtime.zh.md)
|
||||
|
||||
## Problem
|
||||
|
||||
Background bash originally combined two responsibilities: the bash executor ran processes and also managed task ids, ownership, incremental reads, cancellation, completion listeners, and model-facing control tools. Adding background subagents required the same lifecycle and interaction contract. Implementing that contract independently for every long-running capability would duplicate isolation, cleanup, notification, and prompt behavior while teaching the model a different collect-and-stop protocol for each producer.
|
||||
Background bash originally combined two responsibilities: the bash executor ran processes and also managed job ids, ownership, incremental reads, cancellation, completion listeners, and model-facing control tools. Adding background subagents required the same lifecycle and interaction contract. Implementing that contract independently for every long-running capability would duplicate isolation, cleanup, notification, and prompt behavior while teaching the model a different collect-and-stop protocol for each producer.
|
||||
|
||||
The task registry, control tools, and completion notices form one harness capability. Bash and subagents should supply execution-specific hooks without owning generic task behavior.
|
||||
The job registry, control tools, and completion notices form one harness capability. Bash and subagents should supply execution-specific hooks without owning generic task behavior.
|
||||
|
||||
## Decision
|
||||
|
||||
The `tasks/` package group owns background-task semantics:
|
||||
The `jobs/` package group owns background-job semantics:
|
||||
|
||||
- `@deepseek-ai/dsh-tasks` registers running work as `ctx.tasks` and owns task ids, authorization, snapshots, reads, cancellation, waiting, completion listeners, and cleanup.
|
||||
- `@deepseek-ai/dsh-tool-tasks` exposes `task_output`, `task_list`, and `task_kill`, injects completion notices, and supplies the background-task system-prompt guidance.
|
||||
- `@deepseek-ai/dsh-jobs` registers running work as `ctx.jobs` and owns job ids, authorization, snapshots, reads, cancellation, waiting, completion listeners, and cleanup.
|
||||
- `@deepseek-ai/dsh-tool-jobs` exposes `job_output`, `job_list`, and `job_kill`, injects completion notices, and supplies the background-job system-prompt guidance.
|
||||
|
||||
Long-running tools are producers. `dsh-tool-bash` adapts a `BashProcess` into incremental output and process cancellation; `dsh-tool-subagent` adapts a child run into final output and child disposal. The bash and subagent capability seams remain independent of sessions and the task registry.
|
||||
Long-running tools are producers. `dsh-tool-bash` adapts a `ShellProcess` into incremental output and process cancellation; `dsh-tool-subagent` adapts a child run into final output and child disposal. The bash and subagent capability seams remain independent of sessions and the job registry.
|
||||
|
||||
`TaskService` is the Service Definition in `@deepseek-ai/dsh-tasks`; the process-local provider is `LocalTaskService` in `@deepseek-ai/dsh-tasks-local` (the [task-registry contract Agent Note](2026-07-26-task-registry-seam.md) records that split).
|
||||
`JobRegistry` is the Service Definition in `@deepseek-ai/dsh-jobs`; the process-local provider is `LocalJobRegistry` in `@deepseek-ai/dsh-jobs-local` (the [task-registry contract Agent Note](2026-07-26-job-registry-seam.md) records that split).
|
||||
|
||||
## Runtime contract
|
||||
|
||||
The literal types live on the [tasks subsystem page](../../../../docs/subsystems/tasks.md). A producer calls `ctx.tasks.start()` with a kind, label, optional owning `Agent`, optional positive `outputLimitBytes`, and a `run()` function. The runtime completes all failable preflight work before calling `run()` and invokes it once. After `run()` returns hooks, registration commits without another failable step; a producer cannot start work that lacks a collectable task id.
|
||||
The literal types live on the [tasks subsystem page](../../../../docs/subsystems/jobs.md). A producer calls `ctx.jobs.start()` with a kind, label, optional owning `Agent`, optional positive `outputLimitBytes`, and a `run()` function. The runtime completes all failable preflight work before calling `run()` and invokes it once. After `run()` returns hooks, registration commits without another failable step; a producer cannot start work that lacks a collectable job id.
|
||||
|
||||
The process-local provider also owns bounded admission, whose rationale is recorded in the [bounded background task admission decision](../bug-fix/2026-08-11-bounded-background-task-admission.md). Its positive-safe-integer `maxConcurrentTasksPerOwner` config defaults to `10`; `start()` derives each exact `Agent` object's active count from `running` and `stopping` records, while every unowned task shares one service bucket. Capacity rejection occurs before `run()` and id allocation, and producer `done` settlement is the only event that releases a stopping task's place. The provider does not queue, preempt, or retain a second mutable count.
|
||||
The process-local provider also owns bounded admission, whose rationale is recorded in the [bounded background job admission decision](../bug-fix/2026-08-11-bounded-background-job-admission.md). Its positive-safe-integer `maxConcurrentJobsPerOwner` config defaults to `10`; `start()` derives each exact `Agent` object's active count from `running` and `stopping` records, while every unowned task shares one service bucket. Capacity rejection occurs before `run()` and id allocation, and producer `done` settlement is the only event that releases a stopping task's place. The provider does not queue, preempt, or retain a second mutable count.
|
||||
|
||||
`outputLimitBytes` is producer-owned presentation policy, not a registry buffer. The registry validates and projects it unchanged into `TaskSnapshot`; generic control APIs apply the cap to complete model-facing output after adding their own status or notice metadata. Omitting it preserves the existing controller behavior, so the runtime does not impose a hidden default on unrelated producer families.
|
||||
`outputLimitBytes` is producer-owned presentation policy, not a registry buffer. The registry validates and projects it unchanged into `JobSnapshot`; generic control APIs apply the cap to complete model-facing output after adding their own status or notice metadata. Omitting it preserves the existing controller behavior, so the runtime does not impose a hidden default on unrelated producer families.
|
||||
|
||||
A model-facing producer exposes that committed id in its canonical success value, normally `{ kind: 'background', taskId }`; Native rendering may keep human-readable prose. A pre-aborted background call fails rather than returning a no-op because no task exists to satisfy the promised handle. Once registration publishes the id, cancellation belongs to the task's own controller and the task runtime: later cancellation of the producing tool call must not kill the published task. `task_kill`, owner disposal, and service teardown request cancellation; foreground execution remains coupled to the call's `exec.signal`.
|
||||
A model-facing producer exposes that committed id in its canonical success value, normally `{ kind: 'background', jobId }`; Native rendering may keep human-readable prose. A pre-aborted background call fails rather than returning a no-op because no task exists to satisfy the promised handle. Once registration publishes the id, cancellation belongs to the task's own controller and the job runtime: later cancellation of the producing tool call must not kill the published task. `job_kill`, owner disposal, and service teardown request cancellation; foreground execution remains coupled to the call's `exec.signal`.
|
||||
|
||||
The producer hooks define three responsibilities:
|
||||
|
||||
- `cancel(reason?)` synchronously requests termination, is idempotent, and must cause `done` to settle.
|
||||
- `done` never rejects and settles only after the producer has released the task's resources.
|
||||
- Optional `readOutput()` returns the next consuming output delta. Omitting it declares a final-output task whose terminal result comes from `TaskOutcome.output`.
|
||||
- Optional `readOutput()` returns the next consuming output delta. Omitting it declares a final-output task whose terminal result comes from `JobOutcome.output`.
|
||||
|
||||
Statuses are `running`, `stopping`, `completed`, `killed`, and `failed`. Producer-specific information such as an exit code or stop reason belongs in `detail`; the registry does not interpret it. Task kinds form a merge-extensible string union, and task ids are branded and generated as `<kind>-N`, with a counter per kind.
|
||||
Statuses are `running`, `stopping`, `completed`, `killed`, and `failed`. Producer-specific information such as an exit code or stop reason belongs in `detail`; the registry does not interpret it. Task kinds form a merge-extensible string union, and job ids are branded and generated as `<kind>-N`, with a counter per kind.
|
||||
|
||||
The runtime attaches one continuation to `done`, records the first terminal outcome, resolves waiters, and invokes completion listeners with per-listener error containment. First-wins settlement matters during teardown: if `cancel` throws, the runtime force-fails the record and warns that work may be orphaned rather than waiting forever for a promise that may never settle. A later producer outcome cannot overwrite that diagnosis or notify twice. A `cancel` that returns without eventually settling `done` still blocks teardown because the runtime cannot distinguish it from a slow, valid stop.
|
||||
|
||||
@@ -45,7 +45,7 @@ Task registrations are not effects of the producer tool fiber. Reloading a tool
|
||||
|
||||
## Authorization and owner lifecycle
|
||||
|
||||
Task ids are runtime-global and predictable, so every access is authorized by the registry. `get`, `read`, `wait`, and `kill` accept the calling `Agent`; `list` returns only tasks visible to that caller. An owned task is accessible only to the exact owning session. Unowned tasks are open to non-agent callers and die with the task service.
|
||||
Job ids are runtime-global and predictable, so every access is authorized by the registry. `get`, `read`, `wait`, and `kill` accept the calling `Agent`; `list` returns only tasks visible to that caller. An owned task is accessible only to the exact owning session. Unowned tasks are open to non-agent callers and die with the task service.
|
||||
|
||||
The snapshot stores the owner's branded `SessionId` for authorization, while lifecycle operations retain the exact live `Agent` instance. These identities serve different purposes: session equality grants access, but exact object identity selects cleanup and completion delivery. Reusing an agent or session id cannot redirect an old scope's cleanup or notices to a replacement.
|
||||
|
||||
@@ -55,45 +55,45 @@ For contract-compliant producers, `AgentHandle.dispose()` resolves only after ow
|
||||
|
||||
## Service API
|
||||
|
||||
`TaskService` provides:
|
||||
`JobRegistry` provides:
|
||||
|
||||
- `start(spec)` for preflighted, provider-admitted, atomic registration.
|
||||
- `get(id, caller?)` and `list(caller?)` for non-consuming snapshots.
|
||||
- `read(id, caller?)` for a consuming stream delta or an idempotent final result.
|
||||
- `kill(id, caller?, reason?)` for cancellation.
|
||||
- `wait(id, timeoutMs, caller?, signal?)` for bounded terminal waiting.
|
||||
- `onTaskDone(listener)` for effect-scoped observation with exact-owner delivery and listener containment.
|
||||
- `onJobDone(listener)` for effect-scoped observation with exact-owner delivery and listener containment.
|
||||
- `attachController(name)` for the task-controller availability fence.
|
||||
|
||||
`wait` returns the terminal snapshot when the task settles or the live snapshot when its timeout expires. Aborting a wait cancels only that wait. If settlement has already assigned terminal delivery to the waiter, the terminal snapshot still wins. Waiters unregister synchronously on abort so a same-tick settlement cannot suppress a completion notice on behalf of a reader that receives nothing.
|
||||
|
||||
A producer loaded without any controller would let callers start work they cannot collect or stop. `dsh-tool-tasks` therefore calls `attachController()` for its lifetime, and `start()` fails before producer execution when no controller is attached. This check occurs at start rather than plugin load because sibling plugins may activate concurrently. Custom non-model controllers can attach themselves without teaching the registry tool names.
|
||||
A producer loaded without any controller would let callers start work they cannot collect or stop. `dsh-tool-jobs` therefore calls `attachController()` for its lifetime, and `start()` fails before producer execution when no controller is attached. This check occurs at start rather than plugin load because sibling plugins may activate concurrently. Custom non-model controllers can attach themselves without teaching the registry tool names.
|
||||
|
||||
## Model-facing control API
|
||||
|
||||
`dsh-tool-tasks` registers three kind-independent tools with generic UI cards:
|
||||
`dsh-tool-jobs` registers three kind-independent tools with generic UI cards:
|
||||
|
||||
- `task_output(task_id, wait?, timeout_ms?)` reads output and always appends `[status: ...]`. Stream tasks return only output since the previous read; final-output tasks return their result after settlement. Reads are non-blocking unless `wait: true`, whose timeout is defaulted and capped by plugin config. A wait timeout reports the still-running status and does not stop the task.
|
||||
- `task_list()` returns caller-visible tasks as `<id> [<kind>] <status> — <label>`, or `(no background tasks)`.
|
||||
- `task_kill(task_id, reason?)` requests cancellation immediately. The optional logged reason is forwarded to the producer. Terminal tasks report their existing status; a throwing producer cancel fails the call and leaves the task running.
|
||||
- `job_output(job_id, wait?, timeout_ms?)` reads output and always appends `[status: ...]`. Stream tasks return only output since the previous read; final-output tasks return their result after settlement. Reads are non-blocking unless `wait: true`, whose timeout is defaulted and capped by plugin config. A wait timeout reports the still-running status and does not stop the task.
|
||||
- `job_list()` returns caller-visible tasks as `<id> [<kind>] <status> — <label>`, or `(no background jobs)`.
|
||||
- `job_kill(job_id, reason?)` requests cancellation immediately. The optional logged reason is forwarded to the producer. Terminal tasks report their existing status; a throwing producer cancel fails the call and leaves the task running.
|
||||
|
||||
Stream reads share one task-scoped consuming cursor because the owning model is the intended reader. A UI or multiple independent readers need a separate non-consuming observation API; sharing this cursor would let readers consume one another's output.
|
||||
|
||||
The system prompt tells the model to retain task ids, continue independent work instead of busy-polling or duplicating a running task, collect relevant tasks before its final answer, and kill work that no longer matters. Completion delivers a logged message to the exact owner's session. A busy owner is injected; an idle owner is woken, under the bounded policy the [idle-owner wake decision](../feature/2026-08-11-background-task-completion-wakes-an-idle-owner.md) owns.
|
||||
The system prompt tells the model to retain job ids, continue independent work instead of busy-polling or duplicating a running task, collect relevant tasks before its final answer, and kill work that no longer matters. Completion delivers a logged message to the exact owner's session. A busy owner is injected; an idle owner is woken, under the bounded policy the [idle-owner wake decision](../feature/2026-08-11-background-job-completion-wakes-an-idle-owner.md) owns.
|
||||
|
||||
The runtime marks a terminal task `reported` when a read or wait delivers it, when a live waiter has claimed delivery at settlement, or when the model explicitly kills it. Reported tasks do not inject redundant completion notices. Listener failures are logged independently, do not stop later listeners, and are not awaited by waiters or teardown. When a snapshot carries `outputLimitBytes`, `dsh-tool-tasks` preserves UTF-8 boundaries and reuses an existing producer truncation marker rather than duplicating it. Reads reserve status suffixes and retain the output tail; completion notices reserve the stable `background task <id>` prefix and `task_output` instruction before truncating variable kind, label, status, detail, or the truncation marker itself, so the minimum PTY cap still identifies the task to collect. The task controller resolves the caller-visible producer cap in a prepended pre-execute listener before policy can deny or short-circuit dispatch, then applies it through the task definitions' last-mile `finalizeContent` callback so normalized tool errors, outer pipeline failures, and single-text policy results cannot escape the bound; deliberately structured multi-block policy results retain policy ownership of their shape and size.
|
||||
The runtime marks a terminal task `reported` when a read or wait delivers it, when a live waiter has claimed delivery at settlement, or when the model explicitly kills it. Reported tasks do not inject redundant completion notices. Listener failures are logged independently, do not stop later listeners, and are not awaited by waiters or teardown. When a snapshot carries `outputLimitBytes`, `dsh-tool-jobs` preserves UTF-8 boundaries and reuses an existing producer truncation marker rather than duplicating it. Reads reserve status suffixes and retain the output tail; completion notices reserve the stable `background job <id>` prefix and `job_output` instruction before truncating variable kind, label, status, detail, or the truncation marker itself, so the minimum PTY cap still identifies the task to collect. The job controller resolves the caller-visible producer cap in a prepended pre-execute listener before policy can deny or short-circuit dispatch, then applies it through the task definitions' last-mile `finalizeContent` callback so normalized tool errors, outer pipeline failures, and single-text policy results cannot escape the bound; deliberately structured multi-block policy results retain policy ownership of their shape and size.
|
||||
|
||||
## Producer opt-in
|
||||
|
||||
Each producer owns whether its schema exposes `run_in_background` through defaulted config. `dsh-tool-bash`, `dsh-tool-pty`, and each `dsh-tool-subagent` instance use `enableRunInBackground`, defaulting to true. A disabled instance omits the parameter and also rejects a forced background argument at execution because the generic argument validator permits undeclared keys. Schema omission advertises the capability; the execution check enforces it.
|
||||
Each producer owns whether its schema exposes `run_in_background` through defaulted config. `dsh-tool-bash`, `dsh-tool-terminal`, and each `dsh-tool-subagent` instance use `enableRunInBackground`, defaulting to true. A disabled instance omits the parameter and also rejects a forced background argument at execution because the generic argument validator permits undeclared keys. Schema omission advertises the capability; the execution check enforces it.
|
||||
|
||||
`ctx.tasks` does not rewrite producer schemas. A bundle forwards configuration only for producers it owns. If a background call reaches `start()` without an attached controller, the runtime fence fails before execution.
|
||||
`ctx.jobs` does not rewrite producer schemas. A bundle forwards configuration only for producers it owns. If a background call reaches `start()` without an attached controller, the runtime fence fails before execution.
|
||||
|
||||
## Producer integrations
|
||||
|
||||
The bash seam exposes `resolve`, `run`, and `start`. `start(spec)` returns a `BashProcess` with incremental reads, cancellation, exit facts, and a non-rejecting quiescence promise. The local executor retains live handles only so its own disposal can kill and join processes. Foreground callers continue to use `resolve` and `run` directly.
|
||||
The bash seam exposes `resolve`, `run`, and `start`. `start(spec)` returns a `ShellProcess` with incremental reads, cancellation, exit facts, and a non-rejecting quiescence promise. The local executor retains live handles only so its own disposal can kill and join processes. Foreground callers continue to use `resolve` and `run` directly.
|
||||
|
||||
For background bash, `dsh-tool-bash` registers the calling agent as owner. Its hooks map `kill()` to cancellation, `done` to a completed or killed `TaskOutcome`, and `readOutput()` to the process's bounded incremental output plus spill and sandbox notices. Generic task tools own ids, status lines, listing, waiting, and completion notices.
|
||||
For background bash, `dsh-tool-bash` registers the calling agent as owner. Its hooks map `kill()` to cancellation, `done` to a completed or killed `JobOutcome`, and `readOutput()` to the process's bounded incremental output plus spill and sandbox notices. Generic task tools own ids, status lines, listing, waiting, and completion notices.
|
||||
|
||||
For background subagents, `dsh-tool-subagent` creates a task-owned `AbortController` and begins provider startup inside the task starter. Cancellation aborts the same signal before or after provider publication. `done` awaits both the child result and child disposal, maps completed output to a final result, maps abort to `killed`, and maps other stop reasons or infrastructure failures to `failed`. Intermediate child history remains in the child session and is not exposed through `readOutput()`.
|
||||
|
||||
@@ -105,7 +105,7 @@ Separate bash and subagent output/stop tools duplicate ids, isolation, cleanup,
|
||||
|
||||
### An immediate abstract task-runtime backend
|
||||
|
||||
The current `TaskStart.run()` contract passes in-process callbacks and exact `Agent` objects. A durable backend changes identity, restart, ownership, and observation semantics, so at introduction time the registry stayed one concrete service rather than freezing the wrong boundary. The [task-registry contract Agent Note](2026-07-26-task-registry-seam.md) later separated the contract from the process-local implementation without changing these in-process semantics.
|
||||
The current `JobStart.run()` contract passes in-process callbacks and exact `Agent` objects. A durable backend changes identity, restart, ownership, and observation semantics, so at introduction time the registry stayed one concrete service rather than freezing the wrong boundary. The [task-registry contract Agent Note](2026-07-26-job-registry-seam.md) later separated the contract from the process-local implementation without changing these in-process semantics.
|
||||
|
||||
### Consumer-owned authorization or cleanup events
|
||||
|
||||
@@ -113,9 +113,9 @@ Consumer-owned checks invite inconsistent or missing isolation on each new contr
|
||||
|
||||
### Blocking output or a separate wait tool
|
||||
|
||||
Blocking by default would serialize the parent while background work runs. Waiting without reading would add another model call and schema without returning useful information. `task_output(wait: true)` makes blocking explicit and combines it with result delivery.
|
||||
Blocking by default would serialize the parent while background work runs. Waiting without reading would add another model call and schema without returning useful information. `job_output(wait: true)` makes blocking explicit and combines it with result delivery.
|
||||
|
||||
The wait uses the shared deadline primitives but not the generic tool-timeout policy. A wait timeout is a successful observation that returns `[status: running]`; the generic policy would replace it with a timeout error. No tool-call timeout controls task lifetime after a task id has been returned.
|
||||
The wait uses the shared deadline primitives but not the generic tool-timeout policy. A wait timeout is a successful observation that returns `[status: running]`; the generic policy would replace it with a timeout error. No tool-call timeout controls task lifetime after a job id has been returned.
|
||||
|
||||
### Runtime-owned output sinks
|
||||
|
||||
@@ -127,7 +127,7 @@ Authorization, not unguessability, is the access boundary, and ids do not derive
|
||||
|
||||
## Testing
|
||||
|
||||
Unit coverage pins preflight atomicity, per-kind ids, per-exact-owner and unowned-bucket admission, `stopping` occupancy, terminal release, output-limit validation and projection, complete UTF-8 result bounds, stream and final reads, wait timeout and abort races, cancellation, first-wins settlement, listener containment, notice suppression, owner isolation, stale owner instances, owner cleanup, service teardown, and the no-controller fence. Producer tests cover bash process mapping, subagent startup cancellation, terminal mapping, and disposal. Snapshot coverage pins the control-tool schemas, prompt guidance, and an assembled ACP path where the configured limit rejects a second real background Bash task with a `task_kill` recovery action.
|
||||
Unit coverage pins preflight atomicity, per-kind ids, per-exact-owner and unowned-bucket admission, `stopping` occupancy, terminal release, output-limit validation and projection, complete UTF-8 result bounds, stream and final reads, wait timeout and abort races, cancellation, first-wins settlement, listener containment, notice suppression, owner isolation, stale owner instances, owner cleanup, service teardown, and the no-controller fence. Producer tests cover bash process mapping, subagent startup cancellation, terminal mapping, and disposal. Snapshot coverage pins the control-tool schemas, prompt guidance, and an assembled ACP path where the configured limit rejects a second real background Bash task with a `job_kill` recovery action.
|
||||
|
||||
## Consequences
|
||||
|
||||
|
||||
@@ -1,4 +1,4 @@
|
||||
# Agent Note: 后台任务运行时（`ctx.tasks`）与通用任务控制工具
|
||||
# Agent Note: 后台任务运行时（`ctx.jobs`）与通用任务控制工具
|
||||
|
||||
Status: implemented
|
||||
|
||||
@@ -6,38 +6,38 @@ Status: implemented
|
||||
|
||||
## 问题
|
||||
|
||||
后台 bash 原本兼有两项职责：bash 执行器既运行进程，又管理 task id、所有权、增量读取、取消、完成监听器和面向模型的控制工具。新增后台 subagent 需要相同的生命周期与交互约定。如果每种长时间运行能力都独立实现该约定，就会重复隔离、清理、通知和提示词行为，还会让模型为每种生产方学习不同的收集与停止协议。
|
||||
后台 bash 原本兼有两项职责：bash 执行器既运行进程，又管理 job id、所有权、增量读取、取消、完成监听器和面向模型的控制工具。新增后台 subagent 需要相同的生命周期与交互约定。如果每种长时间运行能力都独立实现该约定，就会重复隔离、清理、通知和提示词行为，还会让模型为每种生产方学习不同的收集与停止协议。
|
||||
|
||||
任务注册表、控制工具与完成通知共同构成一项 harness 能力。bash 和 subagent 只提供执行专属的钩子，不拥有通用任务行为。
|
||||
|
||||
## 决策
|
||||
|
||||
`tasks/` 包组拥有后台任务语义：
|
||||
`jobs/` 包组拥有后台任务语义：
|
||||
|
||||
- `@deepseek-ai/dsh-tasks` 将运行中的工作注册为 `ctx.tasks`，并拥有 task id、授权、快照、读取、取消、等待、完成监听器与清理。
|
||||
- `@deepseek-ai/dsh-tool-tasks` 暴露 `task_output`、`task_list` 和 `task_kill`，注入完成通知，并提供后台任务的系统提示词指导。
|
||||
- `@deepseek-ai/dsh-jobs` 将运行中的工作注册为 `ctx.jobs`，并拥有 job id、授权、快照、读取、取消、等待、完成监听器与清理。
|
||||
- `@deepseek-ai/dsh-tool-jobs` 暴露 `job_output`、`job_list` 和 `job_kill`，注入完成通知，并提供后台任务的系统提示词指导。
|
||||
|
||||
长时间运行工具是生产方。`dsh-tool-bash` 将 `BashProcess` 适配为增量输出与进程取消；`dsh-tool-subagent` 将子运行适配为最终输出与子运行释放。bash 与 subagent 能力 seam 保持独立，不依赖会话或任务注册表。
|
||||
长时间运行工具是生产方。`dsh-tool-bash` 将 `ShellProcess` 适配为增量输出与进程取消；`dsh-tool-subagent` 将子运行适配为最终输出与子运行释放。bash 与 subagent 能力 seam 保持独立，不依赖会话或任务注册表。
|
||||
|
||||
`TaskService` 是 `@deepseek-ai/dsh-tasks` 中的 Service Definition；进程内 Service provider 是 `@deepseek-ai/dsh-tasks-local` 中的 `LocalTaskService`（该拆分记录在[任务注册表约定 Agent Note](2026-07-26-task-registry-seam.md)中）。
|
||||
`JobRegistry` 是 `@deepseek-ai/dsh-jobs` 中的 Service Definition；进程内 Service provider 是 `@deepseek-ai/dsh-jobs-local` 中的 `LocalJobRegistry`（该拆分记录在[任务注册表约定 Agent Note](2026-07-26-job-registry-seam.md)中）。
|
||||
|
||||
## 运行时约定
|
||||
|
||||
字面类型见[任务子系统页面](../../../../docs/subsystems/tasks.md)。生产方调用 `ctx.tasks.start()`，传入 kind、label、可选的所属 `Agent`、可选的正数 `outputLimitBytes` 与一个 `run()` 函数。运行时会在调用 `run()` 前完成所有可能失败的预检工作，并且只调用一次。`run()` 返回钩子后，注册过程不会再执行可能失败的步骤而直接提交；生产方无法启动没有可收集 task id 的工作。
|
||||
字面类型见[任务子系统页面](../../../../docs/subsystems/jobs.md)。生产方调用 `ctx.jobs.start()`，传入 kind、label、可选的所属 `Agent`、可选的正数 `outputLimitBytes` 与一个 `run()` 函数。运行时会在调用 `run()` 前完成所有可能失败的预检工作，并且只调用一次。`run()` 返回钩子后，注册过程不会再执行可能失败的步骤而直接提交；生产方无法启动没有可收集 job id 的工作。
|
||||
|
||||
进程内 Service provider 还拥有有界准入，其理由记录在[有界后台任务准入决策](../bug-fix/2026-08-11-bounded-background-task-admission.md)中。它的 `maxConcurrentTasksPerOwner` 配置必须是正的安全整数，默认值为 `10`；`start()` 从 `running` 与 `stopping` 记录派生每个确切 `Agent` 对象的活动数量，而全部无 owner 任务共享一个服务级桶。容量拒绝发生在 `run()` 与 id 分配之前，处于 stopping 的任务只有在生产方 `done` 结算时才释放名额。Service provider 不排队或抢占任务，也不保留第二份可变计数。
|
||||
进程内 Service provider 还拥有有界准入，其理由记录在[有界后台任务准入决策](../bug-fix/2026-08-11-bounded-background-job-admission.md)中。它的 `maxConcurrentJobsPerOwner` 配置必须是正的安全整数，默认值为 `10`；`start()` 从 `running` 与 `stopping` 记录派生每个确切 `Agent` 对象的活动数量，而全部无 owner 任务共享一个服务级桶。容量拒绝发生在 `run()` 与 id 分配之前，处于 stopping 的任务只有在生产方 `done` 结算时才释放名额。Service provider 不排队或抢占任务，也不保留第二份可变计数。
|
||||
|
||||
`outputLimitBytes` 是生产方拥有的呈现策略，而非注册表缓冲区。注册表校验该值，并将其原样投影到 `TaskSnapshot`；通用任务控制器添加自身的状态或通知元数据后，再将该上限应用于完整的面向模型输出。省略该值时保持现有控制器行为，因此运行时不会向无关的生产方类别施加隐式默认值。
|
||||
`outputLimitBytes` 是生产方拥有的呈现策略，而非注册表缓冲区。注册表校验该值，并将其原样投影到 `JobSnapshot`；通用任务控制器添加自身的状态或通知元数据后，再将该上限应用于完整的面向模型输出。省略该值时保持现有控制器行为，因此运行时不会向无关的生产方类别施加隐式默认值。
|
||||
|
||||
面向模型的生产方会在规范成功值中暴露已提交的 id，通常为 `{ kind: 'background', taskId }`；Native 渲染仍可保留便于人类阅读的行文。预先被中止的后台调用会失败，而不是返回空操作，因为不存在可履行所承诺句柄的任务。一旦注册过程发布 id，取消就归任务自身的控制器与任务运行时所有：随后取消生产工具调用不得终止已发布的任务。`task_kill`、所有者资源释放和服务拆除会请求取消；前台执行仍与调用的 `exec.signal` 耦合。
|
||||
面向模型的生产方会在规范成功值中暴露已提交的 id，通常为 `{ kind: 'background', jobId }`；Native 渲染仍可保留便于人类阅读的行文。预先被中止的后台调用会失败，而不是返回空操作，因为不存在可履行所承诺句柄的任务。一旦注册过程发布 id，取消就归任务自身的控制器与任务运行时所有：随后取消生产工具调用不得终止已发布的任务。`job_kill`、所有者资源释放和服务拆除会请求取消；前台执行仍与调用的 `exec.signal` 耦合。
|
||||
|
||||
生产方钩子定义三项职责：
|
||||
|
||||
- `cancel(reason?)` 同步请求终止，具备幂等性，并且必须使 `done` 完成。
|
||||
- `done` 从不拒绝，并且仅在生产方释放任务资源后完成。
|
||||
- 可选的 `readOutput()` 返回下一个消费式输出增量。省略该钩子即声明这是最终输出任务，其终止结果来自 `TaskOutcome.output`。
|
||||
- 可选的 `readOutput()` 返回下一个消费式输出增量。省略该钩子即声明这是最终输出任务，其终止结果来自 `JobOutcome.output`。
|
||||
|
||||
状态包括 `running`、`stopping`、`completed`、`killed` 和 `failed`。退出码或停止原因等生产方专属信息放在 `detail` 中，注册表不解释这些信息。任务 kind 构成可合并扩展的字符串联合；task id 带品牌，并按 `<kind>-N` 生成，每个 kind 各有一个计数器。
|
||||
状态包括 `running`、`stopping`、`completed`、`killed` 和 `failed`。退出码或停止原因等生产方专属信息放在 `detail` 中，注册表不解释这些信息。任务 kind 构成可合并扩展的字符串联合；job id 带品牌，并按 `<kind>-N` 生成，每个 kind 各有一个计数器。
|
||||
|
||||
运行时为 `done` 附加一个 continuation，记录第一个终止结果、使等待方完成等待，并逐个调用完成监听器，同时隔离每个监听器的错误。首次结果优先的结算在资源销毁期间至关重要：如果 `cancel` 抛出，运行时会强制将记录标为失败，并警告工作可能遗留，而不是永远等待一个可能永不完成的 promise。后续生产方结果不能覆盖该诊断，也不能重复通知。`cancel` 返回后如果最终未使 `done` 完成，仍会阻塞资源销毁，因为运行时无法区分这种情况与缓慢但有效的停止。
|
||||
|
||||
@@ -45,7 +45,7 @@ Status: implemented
|
||||
|
||||
## 授权与所有者生命周期
|
||||
|
||||
task id 在运行时全局可见且可预测，因此注册表会授权每次访问。`get`、`read`、`wait` 和 `kill` 接受调用方 `Agent`；`list` 仅返回该调用方可见的任务。有所有者的任务仅允许对应的确切会话访问。无所有者任务向非 agent 调用方开放，并随任务服务一起终止。
|
||||
job id 在运行时全局可见且可预测，因此注册表会授权每次访问。`get`、`read`、`wait` 和 `kill` 接受调用方 `Agent`；`list` 仅返回该调用方可见的任务。有所有者的任务仅允许对应的确切会话访问。无所有者任务向非 agent 调用方开放，并随任务服务一起终止。
|
||||
|
||||
快照存储所有者的品牌化 `SessionId` 以供授权，生命周期操作则保留确切的存活 `Agent` 实例。这两种身份用途不同：会话相等性授予访问权，精确对象身份决定清理和完成通知的接收方。复用 agent 或会话 id，不能将旧作用域的清理或通知重定向到替代实例。
|
||||
|
||||
@@ -55,45 +55,45 @@ task id 在运行时全局可见且可预测，因此注册表会授权每次访
|
||||
|
||||
## 服务 API
|
||||
|
||||
`TaskService` 提供：
|
||||
`JobRegistry` 提供：
|
||||
|
||||
- `start(spec)`：经过预检与 Service provider 准入的原子注册。
|
||||
- `get(id, caller?)` 和 `list(caller?)`：非消费式快照。
|
||||
- `read(id, caller?)`：消费式流增量或幂等的最终结果。
|
||||
- `kill(id, caller?, reason?)`：取消。
|
||||
- `wait(id, timeoutMs, caller?, signal?)`：有界的终止等待。
|
||||
- `onTaskDone(listener)`：effect 作用域内的观察，具有精确所有者投递和监听器隔离。
|
||||
- `onJobDone(listener)`：effect 作用域内的观察，具有精确所有者投递和监听器隔离。
|
||||
- `attachController(name)`：任务控制器可用性防线。
|
||||
|
||||
`wait` 在任务完成时返回终止快照，在等待超时时返回当前快照。中止一次等待只取消该次等待。如果结算已经将终止投递分配给该等待方，终止快照仍然优先。等待方在中止时同步注销，因此同一 tick 内发生结算时，不会代表一个实际未收到任何内容的读取方压制完成通知。
|
||||
|
||||
如果生产方加载时没有任何任务控制器，调用方就能启动无法收集或停止的工作。因此，`dsh-tool-tasks` 在其整个生命周期内调用 `attachController()`；没有附加控制器时，`start()` 会在生产方开始执行前失败。该检查发生在启动时而非插件加载时，因为兄弟插件可能并发激活。自定义的非模型控制器可以自行附加，无需让注册表了解工具名称。
|
||||
如果生产方加载时没有任何任务控制器，调用方就能启动无法收集或停止的工作。因此，`dsh-tool-jobs` 在其整个生命周期内调用 `attachController()`；没有附加控制器时，`start()` 会在生产方开始执行前失败。该检查发生在启动时而非插件加载时，因为兄弟插件可能并发激活。自定义的非模型控制器可以自行附加，无需让注册表了解工具名称。
|
||||
|
||||
## 面向模型的控制 API
|
||||
|
||||
`dsh-tool-tasks` 注册三个与 kind 无关的工具，并使用通用 UI 卡片：
|
||||
`dsh-tool-jobs` 注册三个与 kind 无关的工具，并使用通用 UI 卡片：
|
||||
|
||||
- `task_output(task_id, wait?, timeout_ms?)` 读取输出，并始终追加 `[status: ...]`。流式任务只返回上次读取以来的输出；最终输出任务在结算后返回结果。除非指定 `wait: true`，否则读取不会阻塞；等待超时由插件配置提供默认值并限定上限。等待超时会报告仍在运行的状态，不会停止任务。
|
||||
- `task_list()` 将调用方可见的任务返回为 `<id> [<kind>] <status> — <label>`，没有任务时返回 `(no background tasks)`。
|
||||
- `task_kill(task_id, reason?)` 立即请求取消。可选的已记录原因会转发给生产方。终止任务报告现有状态；生产方的取消操作若抛出，调用便会失败，任务保持运行。
|
||||
- `job_output(job_id, wait?, timeout_ms?)` 读取输出，并始终追加 `[status: ...]`。流式任务只返回上次读取以来的输出；最终输出任务在结算后返回结果。除非指定 `wait: true`，否则读取不会阻塞；等待超时由插件配置提供默认值并限定上限。等待超时会报告仍在运行的状态，不会停止任务。
|
||||
- `job_list()` 将调用方可见的任务返回为 `<id> [<kind>] <status> — <label>`，没有任务时返回 `(no background jobs)`。
|
||||
- `job_kill(job_id, reason?)` 立即请求取消。可选的已记录原因会转发给生产方。终止任务报告现有状态；生产方的取消操作若抛出，调用便会失败，任务保持运行。
|
||||
|
||||
流式读取共享一个任务作用域内的消费游标，因为所属模型是预期读取方。UI 或多个独立读取方需要单独的非消费式观察 API；共享该游标会让读取方彼此消费对方的输出。
|
||||
|
||||
系统提示词要求模型保留 task id、在后台工作运行时继续处理独立工作而非忙轮询或重复启动同一任务、在给出最终答案前收集相关任务，并终止不再重要的工作。完成时，系统会向确切所有者的会话交付一条已记录的消息：繁忙的所有者走注入，空闲的所有者会被唤醒，其有界策略由[空闲所有者唤醒决策](../feature/2026-08-11-background-task-completion-wakes-an-idle-owner.md)负责。
|
||||
系统提示词要求模型保留 job id、在后台工作运行时继续处理独立工作而非忙轮询或重复启动同一任务、在给出最终答案前收集相关任务，并终止不再重要的工作。完成时，系统会向确切所有者的会话交付一条已记录的消息：繁忙的所有者走注入，空闲的所有者会被唤醒，其有界策略由[空闲所有者唤醒决策](../feature/2026-08-11-background-job-completion-wakes-an-idle-owner.md)负责。
|
||||
|
||||
当读取或等待交付终止任务、尚在等待的等待方在结算时认领了投递，或模型显式终止任务时，运行时将终止任务标为 `reported`。已报告的任务不会注入冗余的完成通知。监听器失败会独立记录，不会阻止后续监听器，也不会被等待方或资源销毁过程等待。当快照携带 `outputLimitBytes` 时，`dsh-tool-tasks` 会保持 UTF-8 边界，并复用生产方已有的截断标记，而不会重复添加。读取会为状态后缀预留空间并保留输出尾部；完成通知会先为稳定的 `background task <id>` 前缀与 `task_output` 指令预留空间，再截断可变的 kind、label、status、detail，乃至截断标记本身，因此 PTY 的最小上限仍能标识需要收集的任务。任务控制器在策略有机会拒绝或短路分发之前，于最先执行的 pre-execute 监听器中解析调用方可见的生产方上限；随后通过任务定义最后一道的 `finalizeContent` 回调应用该上限，使规范化的工具错误、外层流水线失败与单文本策略结果都无法绕过该边界；经特意结构化的多块策略结果仍由策略拥有其形状与大小。
|
||||
当读取或等待交付终止任务、尚在等待的等待方在结算时认领了投递，或模型显式终止任务时，运行时将终止任务标为 `reported`。已报告的任务不会注入冗余的完成通知。监听器失败会独立记录，不会阻止后续监听器，也不会被等待方或资源销毁过程等待。当快照携带 `outputLimitBytes` 时，`dsh-tool-jobs` 会保持 UTF-8 边界，并复用生产方已有的截断标记，而不会重复添加。读取会为状态后缀预留空间并保留输出尾部；完成通知会先为稳定的 `background job <id>` 前缀与 `job_output` 指令预留空间，再截断可变的 kind、label、status、detail，乃至截断标记本身，因此 PTY 的最小上限仍能标识需要收集的任务。任务控制器在策略有机会拒绝或短路分发之前，于最先执行的 pre-execute 监听器中解析调用方可见的生产方上限；随后通过任务定义最后一道的 `finalizeContent` 回调应用该上限，使规范化的工具错误、外层流水线失败与单文本策略结果都无法绕过该边界；经特意结构化的多块策略结果仍由策略拥有其形状与大小。
|
||||
|
||||
## 生产方显式启用
|
||||
|
||||
每个生产方通过带默认值的配置，自行决定其 schema 是否暴露 `run_in_background`。`dsh-tool-bash`、`dsh-tool-pty` 和每个 `dsh-tool-subagent` 实例都使用 `enableRunInBackground`，默认值为 true。禁用的实例会省略该参数；由于通用参数校验器允许未声明的键，它还会在执行时拒绝强制传入的后台参数。省略 schema 用于声明能力不可用；执行检查负责强制该约束。
|
||||
每个生产方通过带默认值的配置，自行决定其 schema 是否暴露 `run_in_background`。`dsh-tool-bash`、`dsh-tool-terminal` 和每个 `dsh-tool-subagent` 实例都使用 `enableRunInBackground`，默认值为 true。禁用的实例会省略该参数；由于通用参数校验器允许未声明的键，它还会在执行时拒绝强制传入的后台参数。省略 schema 用于声明能力不可用；执行检查负责强制该约束。
|
||||
|
||||
`ctx.tasks` 不改写生产方 schema。bundle 只转发其所拥有的生产方的配置。如果后台调用在没有附加控制器的情况下到达 `start()`，运行时防线会在执行前使其失败。
|
||||
`ctx.jobs` 不改写生产方 schema。bundle 只转发其所拥有的生产方的配置。如果后台调用在没有附加控制器的情况下到达 `start()`，运行时防线会在执行前使其失败。
|
||||
|
||||
## 生产方集成
|
||||
|
||||
bash seam 暴露 `resolve`、`run` 和 `start`。`start(spec)` 返回一个 `BashProcess`，提供增量读取、取消、退出事实以及不拒绝的完全停稳 promise。本地执行器只为自身释放时能终止并等待进程而保留活动进程的句柄。前台调用方继续直接使用 `resolve` 和 `run`。
|
||||
bash seam 暴露 `resolve`、`run` 和 `start`。`start(spec)` 返回一个 `ShellProcess`，提供增量读取、取消、退出事实以及不拒绝的完全停稳 promise。本地执行器只为自身释放时能终止并等待进程而保留活动进程的句柄。前台调用方继续直接使用 `resolve` 和 `run`。
|
||||
|
||||
对于后台 bash，`dsh-tool-bash` 将调用方 agent 注册为所有者。其钩子将 `kill()` 映射为取消，将 `done` 映射为 completed 或 killed 的 `TaskOutcome`，并将 `readOutput()` 映射为进程的有界增量输出，以及 spill 与沙箱通知。通用任务工具拥有 id、状态行、列表、等待和完成通知。
|
||||
对于后台 bash，`dsh-tool-bash` 将调用方 agent 注册为所有者。其钩子将 `kill()` 映射为取消，将 `done` 映射为 completed 或 killed 的 `JobOutcome`，并将 `readOutput()` 映射为进程的有界增量输出，以及 spill 与沙箱通知。通用任务工具拥有 id、状态行、列表、等待和完成通知。
|
||||
|
||||
对于后台 subagent，`dsh-tool-subagent` 创建由任务拥有的 `AbortController`，并在任务 starter 内启动提供方。无论提供方发布前后，取消都会中止同一个 signal。`done` 同时等待子运行结果和子运行释放，将已完成输出映射为最终结果，将中止映射为 `killed`，并将其他停止原因或基础设施失败映射为 `failed`。中间子历史保留在子会话中，不通过 `readOutput()` 暴露。
|
||||
|
||||
@@ -105,7 +105,7 @@ bash seam 暴露 `resolve`、`run` 和 `start`。`start(spec)` 返回一个 `Bas
|
||||
|
||||
### 立即抽象任务运行时后端
|
||||
|
||||
当前 `TaskStart.run()` 约定传入进程内回调与确切的 `Agent` 对象。持久化后端会改变身份、重启、所有权与观察语义，因此在引入之时注册表保持为单一具体服务，而非固化错误的边界。[任务注册表约定 Agent Note](2026-07-26-task-registry-seam.md)后来在不改变这些进程内语义的前提下，将约定与进程内实现分离。
|
||||
当前 `JobStart.run()` 约定传入进程内回调与确切的 `Agent` 对象。持久化后端会改变身份、重启、所有权与观察语义，因此在引入之时注册表保持为单一具体服务，而非固化错误的边界。[任务注册表约定 Agent Note](2026-07-26-job-registry-seam.md)后来在不改变这些进程内语义的前提下，将约定与进程内实现分离。
|
||||
|
||||
### 由消费方负责授权或清理事件
|
||||
|
||||
@@ -113,9 +113,9 @@ bash seam 暴露 `resolve`、`run` 和 `start`。`start(spec)` 返回一个 `Bas
|
||||
|
||||
### 阻塞输出或单独的等待工具
|
||||
|
||||
默认阻塞会在后台工作运行时串行化父任务。只等待而不读取会增加一次不返回有用信息的模型调用和 schema。`task_output(wait: true)` 显式表达阻塞，并将其与结果交付合并。
|
||||
默认阻塞会在后台工作运行时串行化父任务。只等待而不读取会增加一次不返回有用信息的模型调用和 schema。`job_output(wait: true)` 显式表达阻塞，并将其与结果交付合并。
|
||||
|
||||
等待使用共享的 deadline 原语，而不使用通用工具超时策略。等待超时是一次成功的观察，会返回 `[status: running]`；通用策略会将它替换为超时错误。任务返回 task id 后，没有任何工具调用超时会控制任务生命周期。
|
||||
等待使用共享的 deadline 原语，而不使用通用工具超时策略。等待超时是一次成功的观察，会返回 `[status: running]`；通用策略会将它替换为超时错误。任务返回 job id 后，没有任何工具调用超时会控制任务生命周期。
|
||||
|
||||
### 由运行时拥有输出接收端
|
||||
|
||||
@@ -127,7 +127,7 @@ bash seam 暴露 `resolve`、`run` 和 `start`。`start(spec)` 返回一个 `Bas
|
||||
|
||||
## 测试
|
||||
|
||||
单元覆盖固定预检原子性、按 kind 分配的 id、按确切 owner 与无 owner 桶执行的准入、`stopping` 占位、终态释放、输出上限的校验与投影、完整结果的 UTF-8 字节上限、流式与最终读取、等待超时与中止竞态、取消、首次结果优先的结算、监听器隔离、通知压制、所有者隔离、陈旧的所有者实例、所有者清理、服务资源销毁和无控制器防线。生产方测试覆盖 bash 进程映射、subagent 启动取消、终止映射与释放。快照覆盖固定控制工具 schema、提示词指导，以及一条组合完整的 ACP 路径：配置上限会拒绝第二个真实后台 Bash 任务，并给出 `task_kill` 恢复动作。
|
||||
单元覆盖固定预检原子性、按 kind 分配的 id、按确切 owner 与无 owner 桶执行的准入、`stopping` 占位、终态释放、输出上限的校验与投影、完整结果的 UTF-8 字节上限、流式与最终读取、等待超时与中止竞态、取消、首次结果优先的结算、监听器隔离、通知压制、所有者隔离、陈旧的所有者实例、所有者清理、服务资源销毁和无控制器防线。生产方测试覆盖 bash 进程映射、subagent 启动取消、终止映射与释放。快照覆盖固定控制工具 schema、提示词指导，以及一条组合完整的 ACP 路径：配置上限会拒绝第二个真实后台 Bash 任务，并给出 `job_kill` 恢复动作。
|
||||
|
||||
## 后果
|
||||
|
||||
|
||||
@@ -2,5 +2,5 @@
|
||||
# side as of the last confirmed-consistent state. Both languages carry equal authority;
|
||||
# after editing either side, bring the other along and re-record with:
|
||||
# pnpm run verify-translation-pairing --write .agents/notes/implemented/architecture/2026-06-21-bounded-llm-request-recovery.md
|
||||
2026-06-21-bounded-llm-request-recovery.md: 985ac3803b8d35ea9275052946dc2cbb33480a9f
|
||||
2026-06-21-bounded-llm-request-recovery.zh.md: f1cf4108501fa10588fa7eb26a1334cb5f06855a
|
||||
2026-06-21-bounded-llm-request-recovery.md: 3a91d77fe9c08fbcc3afdce4dad9a38e288b4723
|
||||
2026-06-21-bounded-llm-request-recovery.zh.md: a6b342b19236251e638242c1e9df6b3d6ed557c0
|
||||
|
||||
@@ -52,7 +52,7 @@ The shared transient-code set is intentionally small: adapter mappings for `RATE
|
||||
|
||||
`@deepseek-ai/dsh-llm-retry` is a function plugin that listens to `agent/request-error`. It introduces no service or new loop branch; the agent-loop package changes only the data carried through its existing failed-step recovery control flow.
|
||||
|
||||
The `agent/request-error` waterfall carries the current `LlmFailure`, an immutable list of prior failures that authorized retry turns in the consecutive recovery sequence, and the serving registration's immutable retry policy. The loop transports but does not interpret that policy, owns the consecutive failure history, and clears it after a successful model request. Normal `dsh-llm-retry` policy counts durable retry records scheduled by the same exact-provider policy, while `dsh-compact-basic` keeps its own context-overflow budget. Alternating transient and context-overflow failures therefore consume their owning finite budgets independently; the maximum request count is one plus the sum of the loaded finite budgets.
|
||||
The `agent/request-error` waterfall carries the current `LlmFailure`, an immutable list of prior failures that authorized retry turns in the consecutive recovery sequence, and the serving registration's immutable retry policy. The loop transports but does not interpret that policy, owns the consecutive failure history, and clears it after a successful model request. Normal `dsh-llm-retry` policy counts durable retry records scheduled by the same exact-provider policy, while `dsh-compaction-basic` keeps its own context-overflow budget. Alternating transient and context-overflow failures therefore consume their owning finite budgets independently; the maximum request count is one plus the sum of the loaded finite budgets.
|
||||
|
||||
The [provider-policy decision](../feature/2026-07-24-provider-retry-policies.md) owns the current configuration shape. Provider adapters register their nested `retryPolicy`; omission uses normal defaults: two transient retries, a 500 millisecond initial delay, a 10 second delay cap, 10 percent jitter, and the five transient codes above. The count and delay bounds match the conservative edge of the inspected implementations: [OpenCode uses two request retries with 500 ms/10 s bounds](https://github.com/anomalyco/opencode/blob/9976269ab1accfc9f9dc98a4a688c516934de422/%70ackages/llm/src/route/executor.ts#L36-L39), [Pi separates three agent-level retries from provider retries and defaults provider retries to zero](https://github.com/earendil-works/pi/blob/3da591ab74ab9ab407e72ed882600b2c851fae21/%70ackages/coding-agent/docs/settings.md#L139-L147), and [Codex uses finite request/stream budgets plus a five-minute idle timeout](https://github.com/openai/codex/blob/0fb559f0f6e231a88ac02ea002d3ecd248e2b515/codex-rs/model-provider-info/src/lib.rs#L25-L33). Ten percent follows [Codex's bounded jitter](https://github.com/openai/codex/blob/0fb559f0f6e231a88ac02ea002d3ecd248e2b515/codex-rs/codex-client/src/retry.rs#L40-L47).
|
||||
|
||||
@@ -114,7 +114,7 @@ If recovery is exhausted, the final failure is stored once on `turn/end.reason`
|
||||
- Each provider adapter validates its nested retry policy at Loader startup, and `ctx.llm` captures it with the route; normal mode delegates ineligible paths and makes at most `maxRetries + 1` provider requests when no other policy applies.
|
||||
- HMR-during-backoff tests prove disposal unregisters the listener, aborts and awaits its captured callbacks, emits no retry decision after disposal, and leaves no timer or promise alive.
|
||||
- Pure unit tests cover transient-code selection, exponential backoff and jitter bounds, valid and over-cap `Retry-After`, exhausted budgets, deterministic timer/random hooks, and abort during backoff.
|
||||
- Real agent-loop tests cover failure before chunks, partial chunks then failure, thrown and in-band failures, retry to success in a new turn, exhaustion to structured `turn/end.reason`, and composition with `dsh-compact-basic` context-overflow recovery.
|
||||
- Real agent-loop tests cover failure before chunks, partial chunks then failure, thrown and in-band failures, retry to success in a new turn, exhaustion to structured `turn/end.reason`, and composition with `dsh-compaction-basic` context-overflow recovery.
|
||||
- The partial-chunk integration test proves failed chunks remain attributed to the failed step, no assistant message or tool side effect is committed for that step, and the successful retry records its own chunk seqs and provider/model route.
|
||||
- The plugin-owned `llm/retry` event is non-surface, survives JSONL and SQLite round trips, is ignored by message derivation, and drives TUI and Web retraction plus scheduled-retry rendering. Client tests cover complete wire validation, clock-independent countdown, cancellation versus completed retry labels, and trajectory attribution; keyless UI snapshots cover Web scheduling and success, a real Web composition test covers partial transport failure through recovery, and ACP automation snapshots confirm that a discarded attempt stays off the wire while the recovered reply is emitted.
|
||||
- Idle-watchdog tests prove the stable signal is rearmed only while `next()` is outstanding, disarmed during consumer think time and in `finally`, and classified separately from a total-call deadline and an earlier caller abort; adapter tests prove the signal stops the underlying request rather than merely detaching it.
|
||||
|
||||
@@ -52,7 +52,7 @@ agent loop（智能体循环）会将终止 finish 的 `LlmFailure` 传给 `agen
|
||||
|
||||
`@deepseek-ai/dsh-llm-retry` 是监听 `agent/request-error` 的函数插件。它不引入服务或新的循环分支；agent-loop 包仅会更改通过现有失败步骤恢复控制流携带的数据。
|
||||
|
||||
`agent/request-error` waterfall 携带当前 `LlmFailure`、在连续恢复序列中授权重试轮次的不可变先前失败列表，以及提供服务的注册项所携带的不可变重试策略。循环只传递而不解释该策略；它拥有连续失败历史，并在模型请求成功后清除。`dsh-llm-retry` 的 normal 策略统计由同一项确切提供方策略安排的持久重试记录，`dsh-compact-basic` 则维护自己的上下文溢出预算。因此，暂时性失败与上下文溢出交替出现时，会各自独立消耗其有限预算；最大请求数等于 1 加上所有已加载有限预算之和。
|
||||
`agent/request-error` waterfall 携带当前 `LlmFailure`、在连续恢复序列中授权重试轮次的不可变先前失败列表，以及提供服务的注册项所携带的不可变重试策略。循环只传递而不解释该策略；它拥有连续失败历史，并在模型请求成功后清除。`dsh-llm-retry` 的 normal 策略统计由同一项确切提供方策略安排的持久重试记录，`dsh-compaction-basic` 则维护自己的上下文溢出预算。因此，暂时性失败与上下文溢出交替出现时，会各自独立消耗其有限预算；最大请求数等于 1 加上所有已加载有限预算之和。
|
||||
|
||||
当前配置形状由[提供方策略决策](../feature/2026-07-24-provider-retry-policies.md)规定。提供方适配器会注册嵌套的 `retryPolicy`；省略时使用 normal 默认值：两次暂时性重试、500 毫秒初始延迟、10 秒延迟上限、10% 抖动，以及上述五个暂时性 code。计数与延迟边界参考了所调查实现中较保守的一端：[OpenCode 使用两次请求重试，延迟边界为 500 毫秒／10 秒](https://github.com/anomalyco/opencode/blob/9976269ab1accfc9f9dc98a4a688c516934de422/%70ackages/llm/src/route/executor.ts#L36-L39)；[Pi 将三次 agent 级重试与提供方重试分开，且提供方重试默认为零](https://github.com/earendil-works/pi/blob/3da591ab74ab9ab407e72ed882600b2c851fae21/%70ackages/coding-agent/docs/settings.md#L139-L147)；[Codex 使用有限请求／流预算以及五分钟空闲超时](https://github.com/openai/codex/blob/0fb559f0f6e231a88ac02ea002d3ecd248e2b515/codex-rs/model-provider-info/src/lib.rs#L25-L33)。10% 抖动参考 [Codex 的有界抖动](https://github.com/openai/codex/blob/0fb559f0f6e231a88ac02ea002d3ecd248e2b515/codex-rs/codex-client/src/retry.rs#L40-L47)。
|
||||
|
||||
@@ -114,7 +114,7 @@ agent-spine 演示组合包加载该插件，因此共享的 stdio/TUI、一次
|
||||
- 每个提供方适配器都在 Loader 启动时验证其嵌套重试策略，`ctx.llm` 则将该策略与路由一同捕获；normal mode 会委托不合格路径，而且在没有其他策略时最多发起 `maxRetries + 1` 次提供方请求。
|
||||
- 退避期间执行 HMR 的测试证明：dispose 过程会注销监听器、中止并等待其捕获的回调，dispose 后不发出重试决策，也不留下存活的定时器或 promise。
|
||||
- 纯单元测试覆盖暂时性 code 选择、指数退避和抖动边界、有效及超出上限的 `Retry-After`、耗尽的预算、确定性定时器／随机数钩子，以及退避期间中止。
|
||||
- 真实 agent-loop 测试覆盖分片前失败、部分分片后失败、抛出及带内失败、在新轮次中重试至成功、耗尽后写入结构化 `turn/end.reason`，以及与 `dsh-compact-basic` 上下文溢出恢复的组合。
|
||||
- 真实 agent-loop 测试覆盖分片前失败、部分分片后失败、抛出及带内失败、在新轮次中重试至成功、耗尽后写入结构化 `turn/end.reason`，以及与 `dsh-compaction-basic` 上下文溢出恢复的组合。
|
||||
- 部分分片集成测试证明：失败分片仍归属于失败步骤，该步骤不会提交 assistant 消息或工具副作用，成功的重试会记录自己的分片 seq 和提供方／模型路由。
|
||||
- 插件拥有的不进入表层的 `llm/retry` 事件可在 JSONL 和 SQLite 往返后保留，被消息派生忽略，并驱动 TUI 和 Web 撤回及计划重试渲染。客户端测试覆盖完整的 wire 验证、独立于时钟的倒计时、已取消与已完成重试标签的区别以及轨迹归属；无密钥 UI 快照覆盖 Web 的调度与成功，真实 Web 组合测试覆盖部分传输失败直至恢复，ACP 自动化快照确认，被丢弃的尝试不会通过协议发出，而恢复后的回复会正常发出。
|
||||
- 空闲看门狗测试证明：只有 `next()` 尚未完成时才会重新布防稳定信号；在消费方思考期间及 `finally` 中会解除布防；它与总调用 deadline 以及更早发生的调用方中止分开分类。适配器测试证明该信号会终止底层请求，而不只是与其脱离。
|
||||
|
||||
@@ -2,5 +2,5 @@
|
||||
# side as of the last confirmed-consistent state. Both languages carry equal authority;
|
||||
# after editing either side, bring the other along and re-record with:
|
||||
# pnpm run verify-translation-pairing --write .agents/notes/implemented/architecture/2026-06-21-mandatory-app-attribution-headers.md
|
||||
2026-06-21-mandatory-app-attribution-headers.md: c90b136a556da81e1c15828926f18b865cea4ece
|
||||
2026-06-21-mandatory-app-attribution-headers.zh.md: eb4de8c936f1a83982ff898bc5aa1550b205a572
|
||||
2026-06-21-mandatory-app-attribution-headers.md: 479d3a46dc41c5cc9ae9b77b81dbef3d6524370b
|
||||
2026-06-21-mandatory-app-attribution-headers.zh.md: 75162604623099d50ca87aafd5d45567e2121789
|
||||
|
||||
@@ -71,7 +71,7 @@ The landed contract:
|
||||
|
||||
**Config-only opt-in attribution.** Rejected. A default-off setting is exactly how adapters keep drifting. The policy is mandatory default attribution with overrideable public values, not optional attribution.
|
||||
|
||||
**Product-named token (`deepseek-harness-sdk`).** Considered for the `User-Agent` token, since the product name is DeepSeek Harness SDK. `deepseek-harness` won on continuity: it is the identity providers already see from this codebase, it matches the org/repo identity and package scope, and it keeps wire attribution stable while display copy carries the product name.
|
||||
**SDK-named token (`deepseek-harness-sdk`).** Considered for the `User-Agent` token because the supported runtime client stack uses the SDK name. `deepseek-harness` won because it names the DeepSeek Harness product, matches the org/repo identity and package scope, and keeps wire attribution stable without calling the complete product an SDK.
|
||||
|
||||
## Consequences
|
||||
|
||||
|
||||
@@ -71,7 +71,7 @@ OpenRouter 应用归属刻意未实现。`HTTP-Referer`、`X-OpenRouter-Title`
|
||||
|
||||
**仅配置启用的归属。** 否决。默认关闭的设置正是适配器不断漂移的原因。策略是强制默认归属加可覆盖的公开值，而非可选归属。
|
||||
|
||||
**以产品命名的 token（`deepseek-harness-sdk`）。** 曾考虑用于 `User-Agent` token，因为产品名是 DeepSeek Harness SDK。`deepseek-harness` 因连续性胜出：它是提供方从本代码库已经看到的身份，与组织/仓库身份和包 scope 一致，且在展示文案承载产品名的同时保持线路归属稳定。
|
||||
**以 SDK 命名的 token（`deepseek-harness-sdk`）。** 曾考虑用于 `User-Agent` token，因为受支持的运行时客户端栈使用 SDK 名称。`deepseek-harness` 胜出，因为它命名 DeepSeek Harness 产品、与组织／仓库身份和包 scope 一致，并且在不把完整产品称为 SDK 的前提下保持线路归属稳定。
|
||||
|
||||
## 后果
|
||||
|
||||
|
||||
@@ -2,5 +2,5 @@
|
||||
# side as of the last confirmed-consistent state. Both languages carry equal authority;
|
||||
# after editing either side, bring the other along and re-record with:
|
||||
# pnpm run verify-translation-pairing --write .agents/notes/implemented/architecture/2026-06-24-web-capability-seam.md
|
||||
2026-06-24-web-capability-seam.md: 5df68ea1f0c32491f48d6da98a8fbce0e658e102
|
||||
2026-06-24-web-capability-seam.zh.md: 16ed229ee5d937a1d2a147e5a0693a5e2def0ad5
|
||||
2026-06-24-web-capability-seam.md: 6e81650e2d7286fc3c9df1e740ba9ccbf30118df
|
||||
2026-06-24-web-capability-seam.zh.md: b91a4470999fd66ec7178326fcfb5247ff83a2e0
|
||||
|
||||
@@ -19,7 +19,7 @@ There is also a provider-selection question. Existing `tool-bash` and `tool-fs`
|
||||
Web access is a first-class capability seam following [the capability-seam Agent Note](2026-06-13-capability-seams.md):
|
||||
|
||||
1. `@deepseek-ai/dsh-web` (`packages/web/web`) owns `ctx.web`, provider registration, provider selection, shared request/result vocabulary, and web-specific errors.
|
||||
2. Provider packages implement concrete backends and register capabilities with `ctx.web`, for example `@deepseek-ai/dsh-web-search-exa`, `@deepseek-ai/dsh-web-search-perplexity`, `@deepseek-ai/dsh-web-search-deepseek`, and `@deepseek-ai/dsh-web-fetch-local`.
|
||||
2. Provider packages implement concrete backends and register capabilities with `ctx.web`, for example `@deepseek-ai/dsh-web-search-exa`, `@deepseek-ai/dsh-web-search-perplexity`, `@deepseek-ai/dsh-web-search-deepseek`, and `@deepseek-ai/dsh-web-fetch-http`.
|
||||
3. `@deepseek-ai/dsh-tool-web` (`packages/web/tool-web`) owns the model-facing `web_search` and `web_fetch` tool schemas, prompt sections, argument validation, result formatting, and tool-owned presentation over `ctx.web`.
|
||||
|
||||
Providers do not register tools. Providers register capabilities. `dsh-tool-web` is the only owner of model-facing names, descriptions, prompt guidance, JSON schemas, and presentation.
|
||||
@@ -38,7 +38,7 @@ The seam deliberately exposes no observation surface — no registry-change even
|
||||
|
||||
## Package topology
|
||||
|
||||
The three-package Service Definition / Service provider / Consumer split follows bash and filesystem, but the *interface* package is closer to the LLM seam. `LlmService` (`packages/llm/llm/src/index.ts`) is a name-keyed provider registry: `registerAdapter(models, adapter)` stores adapters in a `Map`, returns a disposer, throws `DUPLICATE_ADAPTER` on duplicate keys, and throws `NO_ADAPTER` at resolution time. `ctx.web` follows that registry shape, but has two capability kinds and a richer selection policy (a configured provider id, or auto-select when exactly one usable provider is registered), so the `WebError` an execution throws can explain why a search or fetch capability cannot run.
|
||||
The three-package Service Definition / Service provider / Consumer split follows bash and filesystem, but the *interface* package is closer to the LLM seam. `LlmRuntime` (`packages/llm/llm/src/index.ts`) is a name-keyed provider registry: `registerAdapter(models, adapter)` stores adapters in a `Map`, returns a disposer, throws `DUPLICATE_ADAPTER` on duplicate keys, and throws `NO_ADAPTER` at resolution time. `ctx.web` follows that registry shape, but has two capability kinds and a richer selection policy (a configured provider id, or auto-select when exactly one usable provider is registered), so the `WebError` an execution throws can explain why a search or fetch capability cannot run.
|
||||
|
||||
The dependency direction mirrors bash and filesystem:
|
||||
|
||||
@@ -49,7 +49,7 @@ The dependency direction mirrors bash and filesystem:
|
||||
implementation
|
||||
<--depends on-- @deepseek-ai/dsh-web-search-deepseek
|
||||
implementation
|
||||
<--depends on-- @deepseek-ai/dsh-web-fetch-local
|
||||
<--depends on-- @deepseek-ai/dsh-web-fetch-http
|
||||
implementation
|
||||
```
|
||||
|
||||
@@ -60,7 +60,7 @@ flowchart LR
|
||||
exa["@deepseek-ai/dsh-web-search-exa"] -->|registerSearchProvider| web["@deepseek-ai/dsh-web / ctx.web"]
|
||||
perplexity["@deepseek-ai/dsh-web-search-perplexity"] -->|registerSearchProvider| web
|
||||
deepseek["@deepseek-ai/dsh-web-search-deepseek"] -->|registerSearchProvider| web
|
||||
fetchLocal["@deepseek-ai/dsh-web-fetch-local"] -->|registerFetchProvider| web
|
||||
fetchLocal["@deepseek-ai/dsh-web-fetch-http"] -->|registerFetchProvider| web
|
||||
toolWeb["@deepseek-ai/dsh-tool-web"] -->|search/fetch| web
|
||||
toolWeb -->|ctx.tools.register| webSearch["tool: web_search"]
|
||||
toolWeb -->|ctx.tools.register| webFetch["tool: web_fetch"]
|
||||
@@ -74,7 +74,7 @@ Provider packages depend only on `dsh-web` and Cordis. They own credentials, end
|
||||
|
||||
## `ctx.web` contract
|
||||
|
||||
`ctx.web` is a provider registry plus a provider-selecting execution API. The registry half stays close to `LlmService`: a `Map<id, provider>` per capability kind, `registerSearchProvider` / `registerFetchProvider` methods that return disposers, duplicate ids that throw `WebError`, and execution-time resolution that throws when the selected provider is absent or unusable. The authoritative signatures live in `packages/web/web/src/types.ts`; the seam's shape:
|
||||
`ctx.web` is a provider registry plus a provider-selecting execution API. The registry half stays close to `LlmRuntime`: a `Map<id, provider>` per capability kind, `registerSearchProvider` / `registerFetchProvider` methods that return disposers, duplicate ids that throw `WebError`, and execution-time resolution that throws when the selected provider is absent or unusable. The authoritative signatures live in `packages/web/web/src/types.ts`; the seam's shape:
|
||||
|
||||
```ts
|
||||
import type { WebFetchRequest, WebFetchResult, WebSearchRequest, WebSearchResult } from '@deepseek-ai/dsh-web'
|
||||
@@ -91,7 +91,7 @@ interface WebFetchProvider {
|
||||
fetch(request: WebFetchRequest, signal?: AbortSignal): Promise<WebFetchResult>
|
||||
}
|
||||
|
||||
interface WebService {
|
||||
interface WebRuntime {
|
||||
registerSearchProvider(provider: WebSearchProvider): () => void
|
||||
registerFetchProvider(provider: WebFetchProvider): () => void
|
||||
|
||||
@@ -108,7 +108,7 @@ Provider ids are stable strings and unique within their capability kind. Registe
|
||||
|
||||
Provider availability and capability selection are separate concepts, but both stay minimal. A provider reports only whether that concrete implementation is usable by cheap local checks such as credential presence or parseable endpoint config. A provider `available()` must not make network calls.
|
||||
|
||||
`LlmService` has no status type at all: availability is expressed as registry membership plus a resolution-time throw. `ctx.web` follows the same discipline. The seam exposes no aggregated capability-status query — `search()` / `fetch()` derive the selection on each call from the configured provider id, the registered providers, and each provider's cheap local `available()` boolean, and a selection failure is the structured `WebError` thrown at execution time. A caller that needs to know whether a capability can run executes and routes that error; nothing is stored as mutable service state.
|
||||
`LlmRuntime` has no status type at all: availability is expressed as registry membership plus a resolution-time throw. `ctx.web` follows the same discipline. The seam exposes no aggregated capability-status query — `search()` / `fetch()` derive the selection on each call from the configured provider id, the registered providers, and each provider's cheap local `available()` boolean, and a selection failure is the structured `WebError` thrown at execution time. A caller that needs to know whether a capability can run executes and routes that error; nothing is stored as mutable service state.
|
||||
|
||||
The boolean is an input to selection, not a health system. `tool-web` never calls a provider's `available()` directly — its only path into the seam is `search()` / `fetch()` — so selection policy has one owner.
|
||||
|
||||
@@ -131,7 +131,7 @@ The "single provider auto-selects" rule is for tests, demos, and simple deployme
|
||||
name: '@deepseek-ai/dsh-web'
|
||||
config:
|
||||
searchProvider: exa
|
||||
fetchProvider: local-http
|
||||
fetchProvider: http
|
||||
|
||||
- id: web-search-exa
|
||||
name: '@deepseek-ai/dsh-web-search-exa'
|
||||
@@ -142,8 +142,8 @@ The "single provider auto-selects" rule is for tests, demos, and simple deployme
|
||||
- id: web-search-deepseek
|
||||
name: '@deepseek-ai/dsh-web-search-deepseek'
|
||||
|
||||
- id: web-fetch-local
|
||||
name: '@deepseek-ai/dsh-web-fetch-local'
|
||||
- id: web-fetch-http
|
||||
name: '@deepseek-ai/dsh-web-fetch-http'
|
||||
|
||||
- id: tool-web
|
||||
name: '@deepseek-ai/dsh-tool-web'
|
||||
@@ -199,7 +199,7 @@ Full page retrieval remains the job of `web_fetch(url)`. Search snippets are dis
|
||||
|
||||
## Fetch request and result schema
|
||||
|
||||
The `web_fetch` implementation is an anonymous public HTTP(S) fetch provider, `local-http`. It fetches bytes from a concrete URL, applies the basic transport hygiene below (http/https-only, credential rejection, byte/time caps, cross-origin redirect blocking), decodes textual content, and returns only the minimal model-useful result: final URL, status code, body, and truncation. It carries no browser cookies, editor credentials, git credentials, internal auth tokens, or implicit access to private services. (Full SSRF / private-network blocking is deferred — see [Deferred work](#deferred-work).)
|
||||
The `web_fetch` implementation is an anonymous public HTTP(S) fetch provider, `http`. It fetches bytes from a concrete URL, applies the basic transport hygiene below (http/https-only, credential rejection, byte/time caps, cross-origin redirect blocking), decodes textual content, and returns only the minimal model-useful result: final URL, status code, body, and truncation. It carries no browser cookies, editor credentials, git credentials, internal auth tokens, or implicit access to private services. (Full SSRF / private-network blocking is deferred — see [Deferred work](#deferred-work).)
|
||||
|
||||
The seam request stays smaller than OpenCode's model-facing tool:
|
||||
|
||||
@@ -274,13 +274,13 @@ The model-facing output is text-first because tool results are `ContentBlock[]`,
|
||||
- `WEB_UNSUPPORTED_CONTENT_TYPE`
|
||||
- `WEB_PROVIDER_ERROR`
|
||||
|
||||
`WEB_DUPLICATE_PROVIDER` is thrown synchronously from `registerSearchProvider` / `registerFetchProvider` when an id is already registered for that capability kind (the analogue of `LlmService`'s `DUPLICATE_ADAPTER`); it is a registration-time programming error, not an execution outcome, but shares the `WebError` code space so callers see one taxonomy. `WEB_PROVIDER_ERROR` is the catch-all for a provider's own failure surfaced through the seam, including network/transport failure in `web-fetch-local` (DNS, connection refused, TLS); there is deliberately no separate `WEB_NETWORK` code — the provider sets a descriptive message so the model and logs can tell a network failure from a provider API failure.
|
||||
`WEB_DUPLICATE_PROVIDER` is thrown synchronously from `registerSearchProvider` / `registerFetchProvider` when an id is already registered for that capability kind (the analogue of `LlmRuntime`'s `DUPLICATE_ADAPTER`); it is a registration-time programming error, not an execution outcome, but shares the `WebError` code space so callers see one taxonomy. `WEB_PROVIDER_ERROR` is the catch-all for a provider's own failure surfaced through the seam, including network/transport failure in `web-fetch-http` (DNS, connection refused, TLS); there is deliberately no separate `WEB_NETWORK` code — the provider sets a descriptive message so the model and logs can tell a network failure from a provider API failure.
|
||||
|
||||
Tool execution lets these errors flow through `ToolRegistry.execute()`, which already converts `HarnessError` into an error tool result with structured metadata. The model gets a readable error message; hooks, tests, and UI code can route on the stable code.
|
||||
Tool execution lets these errors flow through `ToolRuntime.execute()`, which already converts `HarnessError` into an error tool result with structured metadata. The model gets a readable error message; hooks, tests, and UI code can route on the stable code.
|
||||
|
||||
## Testing
|
||||
|
||||
Each layer is pinned at its own boundary: the registry/selection/truncation/abort contract and the `WebError` codes in `dsh-web`; per-provider request/response mapping over recorded fixtures (Perplexity fixtures include URL-only citations so the optional source fields stay honest) plus a self-skipping with-key smoke per real provider; real local-HTTP behavior in `web-fetch-local`; and enablement-driven registration, structured execution errors, and result formatting through the real tool registry in `dsh-tool-web`. A real-Loader smoke guards the two export shapes ([postmortem 0001](../../../../docs/postmortem/0001-acp-default-export-drops-inject.md)): `dsh-web` is a default-exported service, while the providers and `tool-web` are namespace plugins where a stray `export default` would drop `inject`.
|
||||
Each layer is pinned at its own boundary: the registry/selection/truncation/abort contract and the `WebError` codes in `dsh-web`; per-provider request/response mapping over recorded fixtures (Perplexity fixtures include URL-only citations so the optional source fields stay honest) plus a self-skipping with-key smoke per real provider; real local-HTTP behavior in `web-fetch-http`; and enablement-driven registration, structured execution errors, and result formatting through the real tool registry in `dsh-tool-web`. A real-Loader smoke guards the two export shapes ([postmortem 0001](../../../../docs/postmortem/0001-acp-default-export-drops-inject.md)): `dsh-web` is a default-exported service, while the providers and `tool-web` are namespace plugins where a stray `export default` would drop `inject`.
|
||||
|
||||
## Alternatives considered
|
||||
|
||||
@@ -294,7 +294,7 @@ This resembles OpenCode's local web search: one stable `websearch` tool dispatch
|
||||
|
||||
### Split search and fetch into two seams (`dsh-search`, `dsh-fetch`)
|
||||
|
||||
Tempting because the two halves share no request schema and no business logic, so each would map cleanly onto the bash/fs three-package template, and the `Search`/`Fetch` method-pair duplication on `WebService` would disappear. Rejected because the shared machinery — provider-id registry, registration-order-independent selection policy, abort propagation, the `WebError` taxonomy, and the product-facing "how this harness reaches the web" configuration API — is real and would otherwise be duplicated across two near-identical seams. One `ctx.web` middle layer gives the product a single thing to inject and configure and gives provider selection one owner. The price is the parallel `searchX`/`fetchX` method pairs, which is accepted deliberately.
|
||||
Tempting because the two halves share no request schema and no business logic, so each would map cleanly onto the shell/fs three-package template, and the `Search`/`Fetch` method-pair duplication on `WebRuntime` would disappear. Rejected because the shared machinery — provider-id registry, registration-order-independent selection policy, abort propagation, the `WebError` taxonomy, and the product-facing "how this harness reaches the web" configuration API — is real and would otherwise be duplicated across two near-identical seams. One `ctx.web` middle layer gives the product a single thing to inject and configure and gives provider selection one owner. The price is the parallel `searchX`/`fetchX` method pairs, which is accepted deliberately.
|
||||
|
||||
### Choose the first registered provider
|
||||
|
||||
@@ -325,7 +325,7 @@ Rejected for the seam. `prompt` turns fetch into LLM summarization and couples p
|
||||
## Deferred work
|
||||
|
||||
- SSRF / private-network protection for `web_fetch`: block private, loopback, link-local, multicast, and otherwise non-public destinations so `web_fetch` is not an SSRF primitive. Doing it correctly is more than a URL-string check — it needs DNS-resolve-then-connect-to-the-validated-IP (to defeat DNS rebinding / TOCTOU), per-hop re-validation across redirects, and IPv6 edge handling (private ranges, IPv4-mapped addresses). Neither reference implementation surveyed does IP-level blocking (OpenCode does a prefix check then fetches; Claude Code relies on a centralized hostname blocklist plus a "private URLs will fail" prompt), so there is no implementation to copy and this is the harness's only SSRF defense — it warrants its own focused design/spike. Until it lands, `web_fetch` must only be enabled in deployments that cannot reach sensitive internal targets.
|
||||
- A `pdf` `WebFetchBody` kind: the `local-http` provider decodes text-extractable PDFs (best-effort, capped, `truncated`) into a `{ kind: 'pdf'; content; pageCount? }` arm, and `tool-web` renders it. This is fetch, not `web_extract` — PDF retrieval is a concrete HTTP 200 plus deterministic local decoding, not provider-side extraction of a non-HTTP resource. Adding it is a coordinated change across `dsh-web` (declare the arm), the provider (decode + narrow "binary rejection" to "reject binary except text-extractable PDF"; scanned/image PDFs needing OCR stay out of scope), and `tool-web` (render). The closed `WebFetchBody` union makes the consumer side fail to compile until the new arm is handled.
|
||||
- A `pdf` `WebFetchBody` kind: the `http` provider decodes text-extractable PDFs (best-effort, capped, `truncated`) into a `{ kind: 'pdf'; content; pageCount? }` arm, and `tool-web` renders it. This is fetch, not `web_extract` — PDF retrieval is a concrete HTTP 200 plus deterministic local decoding, not provider-side extraction of a non-HTTP resource. Adding it is a coordinated change across `dsh-web` (declare the arm), the provider (decode + narrow "binary rejection" to "reject binary except text-extractable PDF"; scanned/image PDFs needing OCR stay out of scope), and `tool-web` (render). The closed `WebFetchBody` union makes the consumer side fail to compile until the new arm is handled.
|
||||
- Provider-backed extraction as a separate `web_extract` capability, rather than widening `web_fetch` silently.
|
||||
- Permission policy integration: the permission system now exists ([sandbox and approval](../feature/2026-07-06-sandbox.md), [web permission presets](../feature/2026-07-23-web-permission-and-approval.md)) but bundles only sandbox mode and approval policy; web permission policy remains unintegrated.
|
||||
- Provider-neutral search controls beyond `query` and `maxResults`, once Exa and Perplexity can both honor them honestly.
|
||||
|
||||
@@ -19,7 +19,7 @@ harness 需要面向模型的 web 工具，但不能将模型约定绑定到某
|
||||
Web 访问是一个一等能力 seam，遵循[能力 seam Agent Note](2026-06-13-capability-seams.md)：
|
||||
|
||||
1. `@deepseek-ai/dsh-web`（`packages/web/web`）拥有 `ctx.web`、提供方注册、提供方选择、共享的请求/结果词汇，以及 web 特有的错误。
|
||||
2. 提供方包实现具体后端并向 `ctx.web` 注册能力，例如 `@deepseek-ai/dsh-web-search-exa`、`@deepseek-ai/dsh-web-search-perplexity`、`@deepseek-ai/dsh-web-search-deepseek` 和 `@deepseek-ai/dsh-web-fetch-local`。
|
||||
2. 提供方包实现具体后端并向 `ctx.web` 注册能力，例如 `@deepseek-ai/dsh-web-search-exa`、`@deepseek-ai/dsh-web-search-perplexity`、`@deepseek-ai/dsh-web-search-deepseek` 和 `@deepseek-ai/dsh-web-fetch-http`。
|
||||
3. `@deepseek-ai/dsh-tool-web`（`packages/web/tool-web`）拥有面向模型的 `web_search` 和 `web_fetch` 工具 schema、提示词段落、参数校验、结果格式化，以及通过 `ctx.web` 实现的工具展示。
|
||||
|
||||
提供方不注册工具。提供方注册能力。`dsh-tool-web` 是面向模型的名称、描述、提示词引导、JSON Schema、展示的唯一所有者。
|
||||
@@ -38,7 +38,7 @@ Web 访问是一个一等能力 seam，遵循[能力 seam Agent Note](2026-06-13
|
||||
|
||||
## 包拓扑
|
||||
|
||||
由三个包构成的 Service Definition / Service provider / Consumer 拆分沿用 bash 和 filesystem 的模式，但*接口*包更接近 LLM（大语言模型） seam。`LlmService`（`packages/llm/llm/src/index.ts`）是一个按名称键控的提供方注册表：`registerAdapter(models, adapter)` 将适配器存入 `Map`、返回 disposer、对重复键抛出 `DUPLICATE_ADAPTER`、在解析时抛出 `NO_ADAPTER`。`ctx.web` 沿用该注册表形状，但有两种能力类别和更丰富的选择策略（配置的提供方 id，或在恰好只有一个可用提供方注册时自动选择），因此执行时抛出的 `WebError` 能解释搜索或 fetch 能力为何无法运行。
|
||||
由三个包构成的 Service Definition / Service provider / Consumer 拆分沿用 bash 和 filesystem 的模式，但*接口*包更接近 LLM（大语言模型） seam。`LlmRuntime`（`packages/llm/llm/src/index.ts`）是一个按名称键控的提供方注册表：`registerAdapter(models, adapter)` 将适配器存入 `Map`、返回 disposer、对重复键抛出 `DUPLICATE_ADAPTER`、在解析时抛出 `NO_ADAPTER`。`ctx.web` 沿用该注册表形状，但有两种能力类别和更丰富的选择策略（配置的提供方 id，或在恰好只有一个可用提供方注册时自动选择），因此执行时抛出的 `WebError` 能解释搜索或 fetch 能力为何无法运行。
|
||||
|
||||
依赖方向与 bash 和 filesystem 一致：
|
||||
|
||||
@@ -49,7 +49,7 @@ Web 访问是一个一等能力 seam，遵循[能力 seam Agent Note](2026-06-13
|
||||
implementation
|
||||
<--depends on-- @deepseek-ai/dsh-web-search-deepseek
|
||||
implementation
|
||||
<--depends on-- @deepseek-ai/dsh-web-fetch-local
|
||||
<--depends on-- @deepseek-ai/dsh-web-fetch-http
|
||||
implementation
|
||||
```
|
||||
|
||||
@@ -60,7 +60,7 @@ flowchart LR
|
||||
exa["@deepseek-ai/dsh-web-search-exa"] -->|registerSearchProvider| web["@deepseek-ai/dsh-web / ctx.web"]
|
||||
perplexity["@deepseek-ai/dsh-web-search-perplexity"] -->|registerSearchProvider| web
|
||||
deepseek["@deepseek-ai/dsh-web-search-deepseek"] -->|registerSearchProvider| web
|
||||
fetchLocal["@deepseek-ai/dsh-web-fetch-local"] -->|registerFetchProvider| web
|
||||
fetchLocal["@deepseek-ai/dsh-web-fetch-http"] -->|registerFetchProvider| web
|
||||
toolWeb["@deepseek-ai/dsh-tool-web"] -->|search/fetch| web
|
||||
toolWeb -->|ctx.tools.register| webSearch["tool: web_search"]
|
||||
toolWeb -->|ctx.tools.register| webFetch["tool: web_fetch"]
|
||||
@@ -74,7 +74,7 @@ flowchart LR
|
||||
|
||||
## `ctx.web` 约定
|
||||
|
||||
`ctx.web` 是一个提供方注册表加上一个带提供方选择的执行 API。注册表部分与 `LlmService` 保持接近：每种能力类别一个 `Map<id, provider>`，`registerSearchProvider`/`registerFetchProvider` 方法返回 disposer，重复 id 抛出 `WebError`，执行时解析在选定提供方缺失或不可用时抛出异常。权威签名见 `packages/web/web/src/types.ts`；seam 的形状：
|
||||
`ctx.web` 是一个提供方注册表加上一个带提供方选择的执行 API。注册表部分与 `LlmRuntime` 保持接近：每种能力类别一个 `Map<id, provider>`，`registerSearchProvider`/`registerFetchProvider` 方法返回 disposer，重复 id 抛出 `WebError`，执行时解析在选定提供方缺失或不可用时抛出异常。权威签名见 `packages/web/web/src/types.ts`；seam 的形状：
|
||||
|
||||
```ts
|
||||
import type { WebFetchRequest, WebFetchResult, WebSearchRequest, WebSearchResult } from '@deepseek-ai/dsh-web'
|
||||
@@ -91,7 +91,7 @@ interface WebFetchProvider {
|
||||
fetch(request: WebFetchRequest, signal?: AbortSignal): Promise<WebFetchResult>
|
||||
}
|
||||
|
||||
interface WebService {
|
||||
interface WebRuntime {
|
||||
registerSearchProvider(provider: WebSearchProvider): () => void
|
||||
registerFetchProvider(provider: WebFetchProvider): () => void
|
||||
|
||||
@@ -108,7 +108,7 @@ interface WebService {
|
||||
|
||||
提供方可用性与能力选择是两个独立概念，但都保持最小化。提供方仅报告该具体实现是否可用，通过廉价的本地检查（如凭证是否存在、端点配置是否可解析）。提供方的 `available()` 禁止发起网络调用。
|
||||
|
||||
`LlmService` 完全没有状态类型：可用性通过注册表成员资格加解析时抛出来表达。`ctx.web` 遵循同样的纪律。seam 不暴露聚合的能力状态查询——`search()`/`fetch()` 在每次调用时根据配置的提供方 id、已注册的提供方和每个提供方廉价的本地 `available()` 布尔值派生选择结果，选择失败就是执行时抛出的结构化 `WebError`。需要知道某项能力能否运行的调用方通过执行并路由该错误来获知；没有任何东西作为可变服务状态存储。
|
||||
`LlmRuntime` 完全没有状态类型：可用性通过注册表成员资格加解析时抛出来表达。`ctx.web` 遵循同样的纪律。seam 不暴露聚合的能力状态查询——`search()`/`fetch()` 在每次调用时根据配置的提供方 id、已注册的提供方和每个提供方廉价的本地 `available()` 布尔值派生选择结果，选择失败就是执行时抛出的结构化 `WebError`。需要知道某项能力能否运行的调用方通过执行并路由该错误来获知；没有任何东西作为可变服务状态存储。
|
||||
|
||||
该布尔值是选择的输入，而非健康系统。`tool-web` 从不直接调用提供方的 `available()`——它进入 seam 的唯一路径是 `search()`/`fetch()`——因此选择策略只有一个所有者。
|
||||
|
||||
@@ -131,7 +131,7 @@ interface WebService {
|
||||
name: '@deepseek-ai/dsh-web'
|
||||
config:
|
||||
searchProvider: exa
|
||||
fetchProvider: local-http
|
||||
fetchProvider: http
|
||||
|
||||
- id: web-search-exa
|
||||
name: '@deepseek-ai/dsh-web-search-exa'
|
||||
@@ -142,8 +142,8 @@ interface WebService {
|
||||
- id: web-search-deepseek
|
||||
name: '@deepseek-ai/dsh-web-search-deepseek'
|
||||
|
||||
- id: web-fetch-local
|
||||
name: '@deepseek-ai/dsh-web-fetch-local'
|
||||
- id: web-fetch-http
|
||||
name: '@deepseek-ai/dsh-web-fetch-http'
|
||||
|
||||
- id: tool-web
|
||||
name: '@deepseek-ai/dsh-tool-web'
|
||||
@@ -199,7 +199,7 @@ Exa 搜索将提供方扁平 `results[]` 的每一项映射为 `WebSearchSource`
|
||||
|
||||
## Fetch 请求与结果 schema
|
||||
|
||||
`web_fetch` 的实现是一个匿名公开 HTTP(S) fetch 提供方 `local-http`。它从具体 URL 获取字节，应用下述基本传输卫生措施（仅 http/https、拒绝 URL 中的凭证、字节/时间上限、跨源重定向阻断），解码文本内容，并仅返回最小的模型可用结果：最终 URL、状态码、正文和截断标志。它不携带浏览器 cookie、编辑器凭证、git 凭证、内部认证令牌，也不隐式访问私有服务。（完整的 SSRF/私有网络阻断推迟——见[推迟工作](#deferred-work)。）
|
||||
`web_fetch` 的实现是一个匿名公开 HTTP(S) fetch 提供方 `http`。它从具体 URL 获取字节，应用下述基本传输卫生措施（仅 http/https、拒绝 URL 中的凭证、字节/时间上限、跨源重定向阻断），解码文本内容，并仅返回最小的模型可用结果：最终 URL、状态码、正文和截断标志。它不携带浏览器 cookie、编辑器凭证、git 凭证、内部认证令牌，也不隐式访问私有服务。（完整的 SSRF/私有网络阻断推迟——见[推迟工作](#deferred-work)。）
|
||||
|
||||
seam 请求比 OpenCode 的面向模型工具更小：
|
||||
|
||||
@@ -274,13 +274,13 @@ SSRF/私有网络防护（阻断私有、回环、链路本地、多播及其他
|
||||
- `WEB_UNSUPPORTED_CONTENT_TYPE`
|
||||
- `WEB_PROVIDER_ERROR`
|
||||
|
||||
`WEB_DUPLICATE_PROVIDER` 在 `registerSearchProvider`/`registerFetchProvider` 发现该能力类别中已有相同 id 时同步抛出（类似 `LlmService` 的 `DUPLICATE_ADAPTER`）；它是注册时的编程错误而非执行结果，但共享 `WebError` 码空间，使调用方看到统一的分类体系。`WEB_PROVIDER_ERROR` 是提供方自身失败通过 seam 浮出的兜底码，包括 `web-fetch-local` 中的网络/传输失败（DNS、连接拒绝、TLS）；刻意不设单独的 `WEB_NETWORK` 码——提供方设置描述性消息，使模型和日志能区分网络失败与提供方 API 失败。
|
||||
`WEB_DUPLICATE_PROVIDER` 在 `registerSearchProvider`/`registerFetchProvider` 发现该能力类别中已有相同 id 时同步抛出（类似 `LlmRuntime` 的 `DUPLICATE_ADAPTER`）；它是注册时的编程错误而非执行结果，但共享 `WebError` 码空间，使调用方看到统一的分类体系。`WEB_PROVIDER_ERROR` 是提供方自身失败通过 seam 浮出的兜底码，包括 `web-fetch-http` 中的网络/传输失败（DNS、连接拒绝、TLS）；刻意不设单独的 `WEB_NETWORK` 码——提供方设置描述性消息，使模型和日志能区分网络失败与提供方 API 失败。
|
||||
|
||||
工具执行让这些错误流经 `ToolRegistry.execute()`，后者已将 `HarnessError` 转换为带结构化元数据的错误工具结果。模型得到可读的错误消息；钩子、测试和 UI 代码可以根据稳定的错误码路由。
|
||||
工具执行让这些错误流经 `ToolRuntime.execute()`，后者已将 `HarnessError` 转换为带结构化元数据的错误工具结果。模型得到可读的错误消息；钩子、测试和 UI 代码可以根据稳定的错误码路由。
|
||||
|
||||
## 测试
|
||||
|
||||
每一层在自己的边界处固定：`dsh-web` 中的注册/选择/截断/abort 约定与 `WebError` 码；每个提供方基于录制的 fixture（测试前置数据）的请求/响应映射（Perplexity fixture 包含纯 URL 引用，以保持可选 source 字段的诚实性），加上每个真实提供方的自跳过带密钥冒烟测试；`web-fetch-local` 中的真实本地 HTTP 行为；`dsh-tool-web` 中通过真实工具注册表的启用驱动注册、结构化执行错误和结果格式化。一个真实 Loader 冒烟测试守护两种导出形状（[事故复盘（postmortem） 0001](../../../../docs/postmortem/0001-acp-default-export-drops-inject.md)）：`dsh-web` 是默认导出的服务，而提供方和 `tool-web` 是命名空间插件，误加 `export default` 会丢失 `inject`。
|
||||
每一层在自己的边界处固定：`dsh-web` 中的注册/选择/截断/abort 约定与 `WebError` 码；每个提供方基于录制的 fixture（测试前置数据）的请求/响应映射（Perplexity fixture 包含纯 URL 引用，以保持可选 source 字段的诚实性），加上每个真实提供方的自跳过带密钥冒烟测试；`web-fetch-http` 中的真实本地 HTTP 行为；`dsh-tool-web` 中通过真实工具注册表的启用驱动注册、结构化执行错误和结果格式化。一个真实 Loader 冒烟测试守护两种导出形状（[事故复盘（postmortem） 0001](../../../../docs/postmortem/0001-acp-default-export-drops-inject.md)）：`dsh-web` 是默认导出的服务，而提供方和 `tool-web` 是命名空间插件，误加 `export default` 会丢失 `inject`。
|
||||
|
||||
## 曾考虑的替代方案
|
||||
|
||||
@@ -294,7 +294,7 @@ SSRF/私有网络防护（阻断私有、回环、链路本地、多播及其他
|
||||
|
||||
### 将搜索和 fetch 拆为两个 seam（`dsh-search`、`dsh-fetch`）
|
||||
|
||||
很有吸引力，因为两半不共享请求 schema 和业务逻辑，各自能干净地映射到 bash/fs 的三包模板上，且 `WebService` 上的 `Search`/`Fetch` 方法对重复也会消失。否决，因为共享的机制——提供方 id 注册表、不依赖注册顺序的选择策略、abort 传播、`WebError` 分类体系，以及面向产品的「这个 harness 如何触达 web」配置 API——是真实存在的，否则会在两个几乎相同的 seam 之间重复。一个 `ctx.web` 中间层给产品一个统一的注入和配置对象，给提供方选择一个唯一的所有者。代价是并行的 `searchX`/`fetchX` 方法对，这是有意接受的。
|
||||
很有吸引力，因为两半不共享请求 schema 和业务逻辑，各自能干净地映射到 shell/fs 的三包模板上，且 `WebRuntime` 上的 `Search`/`Fetch` 方法对重复也会消失。否决，因为共享的机制——提供方 id 注册表、不依赖注册顺序的选择策略、abort 传播、`WebError` 分类体系，以及面向产品的「这个 harness 如何触达 web」配置 API——是真实存在的，否则会在两个几乎相同的 seam 之间重复。一个 `ctx.web` 中间层给产品一个统一的注入和配置对象，给提供方选择一个唯一的所有者。代价是并行的 `searchX`/`fetchX` 方法对，这是有意接受的。
|
||||
|
||||
### 选择第一个注册的提供方
|
||||
|
||||
@@ -327,7 +327,7 @@ SSRF/私有网络防护（阻断私有、回环、链路本地、多播及其他
|
||||
## 推迟工作
|
||||
|
||||
- `web_fetch` 的 SSRF/私有网络防护：阻断私有、回环、链路本地、多播及其他非公开目的地，使 `web_fetch` 不再是 SSRF 原语。正确实现不仅仅是 URL 字符串检查——需要先 DNS 解析再连接到已验证的 IP（防御 DNS rebinding/TOCTOU）、跨重定向的每跳重新验证，以及 IPv6 边缘处理（私有范围、IPv4 映射地址）。所调研的参考实现均未做 IP 级阻断（OpenCode 做前缀检查后直接 fetch；Claude Code 依赖集中式主机名黑名单加「私有 URL 会失败」的提示词），因此没有可复制的实现，且这是 harness 唯一的 SSRF 防线——值得一次专门的设计/spike。在其落地之前，`web_fetch` 只能在无法触达敏感内部目标的部署中启用。
|
||||
- `pdf` `WebFetchBody` 类别：`local-http` 提供方将可文本提取的 PDF 解码（尽力而为、有上限、`truncated`）为 `{ kind: 'pdf'; content; pageCount? }` 分支，`tool-web` 渲染它。这是 fetch 而非 `web_extract`——PDF 获取是具体的 HTTP 200 加确定性的本地解码，不是提供方侧对非 HTTP 资源的提取。添加它是跨 `dsh-web`（声明分支）、提供方（解码 + 将「二进制拒绝」收窄为「拒绝二进制，但可文本提取的 PDF 除外」；需要 OCR 的扫描/图片 PDF 不在范围内）和 `tool-web`（渲染）的协调变更。封闭的 `WebFetchBody` 联合类型使消费方在新分支被处理之前编译失败。
|
||||
- `pdf` `WebFetchBody` 类别：`http` 提供方将可文本提取的 PDF 解码（尽力而为、有上限、`truncated`）为 `{ kind: 'pdf'; content; pageCount? }` 分支，`tool-web` 渲染它。这是 fetch 而非 `web_extract`——PDF 获取是具体的 HTTP 200 加确定性的本地解码，不是提供方侧对非 HTTP 资源的提取。添加它是跨 `dsh-web`（声明分支）、提供方（解码 + 将「二进制拒绝」收窄为「拒绝二进制，但可文本提取的 PDF 除外」；需要 OCR 的扫描/图片 PDF 不在范围内）和 `tool-web`（渲染）的协调变更。封闭的 `WebFetchBody` 联合类型使消费方在新分支被处理之前编译失败。
|
||||
- 提供方支撑的提取作为独立的 `web_extract` 能力，而非静默扩展 `web_fetch`。
|
||||
- 权限策略集成：权限系统现已存在（[沙箱与审批](../feature/2026-07-06-sandbox.md)、[web 权限预设](../feature/2026-07-23-web-permission-and-approval.md)），但只捆绑了沙箱模式与审批策略；web 权限策略仍未集成。
|
||||
- `query` 和 `maxResults` 之外的提供方无关搜索控制，待 Exa 和 Perplexity 都能诚实遵守时再添加。
|
||||
|
||||
@@ -2,5 +2,5 @@
|
||||
# side as of the last confirmed-consistent state. Both languages carry equal authority;
|
||||
# after editing either side, bring the other along and re-record with:
|
||||
# pnpm run verify-translation-pairing --write .agents/notes/implemented/architecture/2026-06-26-file-context-as-event-gate.md
|
||||
2026-06-26-file-context-as-event-gate.md: 5f61201c8129b74233222bdc71eaf20443794760
|
||||
2026-06-26-file-context-as-event-gate.zh.md: 0804e7a1165ea1e2dda05b0b925712e59a63333d
|
||||
2026-06-26-file-context-as-event-gate.md: a8316741c3ce8548db5b5def431425589994995c
|
||||
2026-06-26-file-context-as-event-gate.zh.md: 78b1b759edc0340fd7be103fc9bfec0b125e0800
|
||||
|
||||
@@ -1,4 +1,4 @@
|
||||
# Agent Note: Make `dsh-fs-policy` an event-gate plugin, not a method interface
|
||||
# Agent Note: Make `dsh-fs-observation-policy` an event-gate plugin, not a method interface
|
||||
|
||||
Status: implemented
|
||||
|
||||
@@ -11,19 +11,19 @@ English | [中文](2026-06-26-file-context-as-event-gate.zh.md)
|
||||
This couples three things that should be separable:
|
||||
|
||||
1. **What the tool does** — resolve a path, read a window, write/edit a file. This is the tool's job and needs only `ctx.fs`.
|
||||
2. **The freshness/observation policy** — "edit requires a prior read", "write/edit must be based on the version you read". This is the `dsh-fs-policy` plugin's job.
|
||||
2. **The freshness/observation policy** — "edit requires a prior read", "write/edit must be based on the version you read". This is the `dsh-fs-observation-policy` plugin's job.
|
||||
3. **The recording of observed state** — a side effect that should never block the tool from functioning.
|
||||
|
||||
Because the tool calls `fileContext` methods, removing the policy layer is a breaking change rather than a graceful loss of an *add-on*. The policy is load-bearing for the tool to even run, not an opt-in tightening.
|
||||
|
||||
## Decision
|
||||
|
||||
Invert the control flow. **`dsh-tool-fs` becomes the executor and calls `ctx.fs` directly**; **`dsh-fs-policy` becomes a gate + recorder plugin** that participates through events, never through a method the tool calls and never by registering a `ctx.fileContext` service.
|
||||
Invert the control flow. **`dsh-tool-fs` becomes the executor and calls `ctx.fs` directly**; **`dsh-fs-observation-policy` becomes a gate + recorder plugin** that participates through events, never through a method the tool calls and never by registering a `ctx.fileContext` service.
|
||||
|
||||
```text
|
||||
tool dsh-tool-fs executor: resolves, reads windows, writes/edits via ctx.fs;
|
||||
emits fs policy events; renders results
|
||||
policy dsh-fs-policy plugin: listens to fs/write-intent +
|
||||
policy dsh-fs-observation-policy plugin: listens to fs/write-intent +
|
||||
fs/edit-intent (single-slot waterfall) and fs/observed
|
||||
(emit) events; adds observed-state + freshness.
|
||||
provider contract dsh-fs ctx.fs: text IO + ATOMIC mutation primitives whose version
|
||||
@@ -31,20 +31,20 @@ provider contract dsh-fs ctx.fs: text IO + ATOMIC mutation primitives
|
||||
provider dsh-fs-local local implementation of ctx.fs
|
||||
```
|
||||
|
||||
The model is additive: bare `ctx.fs` performs atomic, unconstrained text I/O, while `dsh-fs-policy` adds observed state, read-before-edit, and version guards. Removing the policy therefore leaves the tools usable but unconstrained. Shipped agent configs load the policy; the bare mode exists to keep policy optional at the service boundary, not as the normal deployment stance.
|
||||
The model is additive: bare `ctx.fs` performs atomic, unconstrained text I/O, while `dsh-fs-observation-policy` adds observed state, read-before-edit, and version guards. Removing the policy therefore leaves the tools usable but unconstrained. Shipped agent configs load the policy; the bare mode exists to keep policy optional at the service boundary, not as the normal deployment stance.
|
||||
|
||||
The [filesystem absence-observation follow-up](../bug-fix/2026-08-09-filesystem-absence-observation.md) refines the recording payload from a success-only version to explicit present/absent state and requires guarded creation to publish without replacement. The event-gate ownership and no-I/O policy boundary remain unchanged.
|
||||
|
||||
`dsh-tool-fs` no longer injects `fileContext`. It injects `fs` and `tools`/`systemPrompt`.
|
||||
|
||||
## The policy is enforced by provider CAS, not by `dsh-fs-policy` stat
|
||||
## The policy is enforced by provider CAS, not by `dsh-fs-observation-policy` stat
|
||||
|
||||
`dsh-fs-policy` enforces "you must write/edit based on the version you read" **without ever calling `stat` or comparing versions itself**. It supplies the observed version as the CAS basis and lets the provider's mutation critical section detect staleness:
|
||||
`dsh-fs-observation-policy` enforces "you must write/edit based on the version you read" **without ever calling `stat` or comparing versions itself**. It supplies the observed version as the CAS basis and lets the provider's mutation critical section detect staleness:
|
||||
|
||||
- "What did this owner last observe?" is the one thing `dsh-fs-policy` decides locally — a `WeakMap` lookup, no I/O. No record means unseen; an absent record permits only guarded creation; a present record carries the replacement/edit basis.
|
||||
- "Is the version still current, or is the create target still absent?" is decided **inside the provider's atomic mutation boundary**. `dsh-fs-policy` supplies `replaceIfVersion` or `createIfAbsent`; the provider raises `FS_STALE_VERSION` for a moved version and `FS_NOT_OBSERVED` when a guarded create loses to another creator.
|
||||
- "What did this owner last observe?" is the one thing `dsh-fs-observation-policy` decides locally — a `WeakMap` lookup, no I/O. No record means unseen; an absent record permits only guarded creation; a present record carries the replacement/edit basis.
|
||||
- "Is the version still current, or is the create target still absent?" is decided **inside the provider's atomic mutation boundary**. `dsh-fs-observation-policy` supplies `replaceIfVersion` or `createIfAbsent`; the provider raises `FS_STALE_VERSION` for a moved version and `FS_NOT_OBSERVED` when a guarded create loses to another creator.
|
||||
|
||||
This is deliberate. If `dsh-fs-policy` stat-ed and compared versions in its waterfall handler, there would be a TOCTOU gap between that check and the tool's actual write — the file could change in between, so the check would be a false guarantee that the provider's lock has to back up anyway. Putting the version check in the provider's critical section is both race-free and zero extra `stat`. So `dsh-fs-policy` does **no** filesystem I/O; the "must be based on the latest read" guarantee is *realized* by CAS, and `dsh-fs-policy` only chooses the basis (`vObserved`) and gates on prior observation.
|
||||
This is deliberate. If `dsh-fs-observation-policy` stat-ed and compared versions in its waterfall handler, there would be a TOCTOU gap between that check and the tool's actual write — the file could change in between, so the check would be a false guarantee that the provider's lock has to back up anyway. Putting the version check in the provider's critical section is both race-free and zero extra `stat`. So `dsh-fs-observation-policy` does **no** filesystem I/O; the "must be based on the latest read" guarantee is *realized* by CAS, and `dsh-fs-observation-policy` only chooses the basis (`vObserved`) and gates on prior observation.
|
||||
|
||||
## Provider contract change: the version guard is optional
|
||||
|
||||
@@ -54,7 +54,7 @@ For the bare provider to be unconstrained, the version guard on its two mutation
|
||||
// writeText: expected is now optional. The FsWriteIntent union is UNCHANGED.
|
||||
writeText(target: FsTarget, content: string, expected?: FsWriteIntent, signal?: AbortSignal): Promise<FsWriteOutcome>
|
||||
// undefined → unconditionally create-or-overwrite (bare default)
|
||||
// createIfAbsent → create only, reject an existing file (dsh-fs-policy, unobserved) [unchanged]
|
||||
// createIfAbsent → create only, reject an existing file (dsh-fs-observation-policy, unobserved) [unchanged]
|
||||
// replaceIfVersion → overwrite only at the observed version, else FS_STALE_VERSION [unchanged]
|
||||
|
||||
// editText: expected becomes optional (was the required { version: FsVersion }).
|
||||
@@ -64,17 +64,17 @@ editText(target: FsTarget, edit: FsEditRequest, expected?: { version: FsVersion
|
||||
// { version } → edit only at that version, else FS_STALE_VERSION (the current behavior)
|
||||
```
|
||||
|
||||
The `FsWriteIntent` union itself does not change — the third "unconditional" state is expressed by *omitting* `expected`, so both mutations share one symmetric shape (`expected?`: omit = no guard, present = guarded). This keeps full backward compatibility for the guarded paths `dsh-fs-policy` uses; only the previously-impossible "no guard" case is new, and it is the bare-provider default. The mutation still runs inside the backend's per-target lock either way, so an unconditional write/edit is still atomic (no torn files); "unconditional" drops the *version* precondition, not the atomicity. `editText` reports a missing target as `FS_STALE_VERSION` on both guarded and unguarded paths, preserving one edit failure code for "the target cannot be edited at this moment".
|
||||
The `FsWriteIntent` union itself does not change — the third "unconditional" state is expressed by *omitting* `expected`, so both mutations share one symmetric shape (`expected?`: omit = no guard, present = guarded). This keeps full backward compatibility for the guarded paths `dsh-fs-observation-policy` uses; only the previously-impossible "no guard" case is new, and it is the bare-provider default. The mutation still runs inside the backend's per-target lock either way, so an unconditional write/edit is still atomic (no torn files); "unconditional" drops the *version* precondition, not the atomicity. `editText` reports a missing target as `FS_STALE_VERSION` on both guarded and unguarded paths, preserving one edit failure code for "the target cannot be edited at this moment".
|
||||
|
||||
## Event vocabulary (owned by `dsh-fs`)
|
||||
|
||||
The events live in `@deepseek-ai/dsh-fs`, not in `dsh-fs-policy`. This is forced by the decoupling contract: `dsh-tool-fs` is the emitter, so it must reference the event types, and it must keep compiling even though `dsh-fs-policy` no longer provides a method service. `dsh-fs` is the package both `dsh-tool-fs` and `dsh-fs-policy` already depend on, so it is the only home that lets the emitter and the policy listener share a vocabulary without the emitter depending on the policy plugin.
|
||||
The events live in `@deepseek-ai/dsh-fs`, not in `dsh-fs-observation-policy`. This is forced by the decoupling contract: `dsh-tool-fs` is the emitter, so it must reference the event types, and it must keep compiling even though `dsh-fs-observation-policy` no longer provides a method service. `dsh-fs` is the package both `dsh-tool-fs` and `dsh-fs-observation-policy` already depend on, so it is the only home that lets the emitter and the policy listener share a vocabulary without the emitter depending on the policy plugin.
|
||||
|
||||
These events carry existing `dsh-fs` vocabulary (`FsTarget`, `FsVersion`, `FsObservation`, `FsWriteIntent`) plus an opaque actor — not model-facing concepts (no line windows, numbered lines, or rendered footers leak down).
|
||||
|
||||
**The two `fs/*` decision events are single-slot, first-wins waterfalls.** `dsh-fs-policy` returns without calling `next()`, so it owns the slot in the default deployment; a listener registered earlier or with `prepend` would replace that policy. Permission, audit, and sandbox concerns remain on the composable `tools/execute` waterfall.
|
||||
**The two `fs/*` decision events are single-slot, first-wins waterfalls.** `dsh-fs-observation-policy` returns without calling `next()`, so it owns the slot in the default deployment; a listener registered earlier or with `prepend` would replace that policy. Permission, audit, and sandbox concerns remain on the composable `tools/execute` waterfall.
|
||||
|
||||
The actor is typed `object` in `dsh-fs` — a pure opaque carrier the provider contract never reads or narrows. The owner-derivation (`actor.agent?.session`) and the `{ agent?: { session? } }` structural shape stay entirely inside `dsh-fs-policy`, which narrows the `object` actor to that shape in its listeners. `dsh-fs` owns the event names and the fs vocabulary; it does NOT own the policy layer's runtime owner structure.
|
||||
The actor is typed `object` in `dsh-fs` — a pure opaque carrier the provider contract never reads or narrows. The owner-derivation (`actor.agent?.session`) and the `{ agent?: { session? } }` structural shape stay entirely inside `dsh-fs-observation-policy`, which narrows the `object` actor to that shape in its listeners. `dsh-fs` owns the event names and the fs vocabulary; it does NOT own the policy layer's runtime owner structure.
|
||||
|
||||
```ts
|
||||
import type { FsObservation, FsTarget, FsVersion, FsWriteIntent } from '@deepseek-ai/dsh-fs'
|
||||
@@ -99,7 +99,7 @@ interface Events {
|
||||
/**
|
||||
* Record that an actor observed a target as present at a version or absent.
|
||||
* Fire-and-forget (plain emit). Listeners MUST be
|
||||
* synchronous, side-effect-only recorders (`dsh-fs-policy`'s is a WeakMap
|
||||
* synchronous, side-effect-only recorders (`dsh-fs-observation-policy`'s is a WeakMap
|
||||
* write); the tool does not guard the emit, so a throwing listener surfaces as
|
||||
* the tool's isError result. No listener ⇒ nothing recorded.
|
||||
* @mode emit
|
||||
@@ -112,7 +112,7 @@ The `fs/*` decision events are **unbound waterfalls dispatched by the tool** (li
|
||||
|
||||
## Tool contract (`dsh-tool-fs`)
|
||||
|
||||
The tool keeps its model-facing schemas (`read`/`write`/`edit`, byte-for-byte unchanged) and prompt sections. The prompt guidance stays policy-first because a deployment loading the fs tools is expected to also load `dsh-fs-policy`: the model is still told to read before overwriting or editing, and that requirement is the fs-policy plugin's, not the backend's. The bare-provider fallback does not change the prompt stance.
|
||||
The tool keeps its model-facing schemas (`read`/`write`/`edit`, byte-for-byte unchanged) and prompt sections. The prompt guidance stays policy-first because a deployment loading the fs tools is expected to also load `dsh-fs-observation-policy`: the model is still told to read before overwriting or editing, and that requirement is the fs-observation-policy plugin's, not the backend's. The bare-provider fallback does not change the prompt stance.
|
||||
|
||||
`dsh-tool-fs` gains the executor responsibilities relocated from the old `fileContext` method service, including **read rendering** (`read-render.ts`: `buildWindow` + `formatReadOutput`, `READ_MAX_BYTES`, `READ_MAX_LINE_LENGTH`, `FileReadOutcome`/`FileTextLine`, plus `STREAM_MIN_SIZE` in `read.ts`), which is the tool's rendering detail now that the tool owns the read. Those read-rendering types and helpers move into `dsh-tool-fs`; the policy plugin must not remain a type dependency for the tool.
|
||||
|
||||
@@ -121,16 +121,16 @@ The tool keeps its model-facing schemas (`read`/`write`/`edit`, byte-for-byte un
|
||||
`stat` budget is minimized by letting the waterfall produce the expectation lazily — the bare default returns `undefined` (no guard) and never stats:
|
||||
|
||||
- **read** — one `stat`; a metadata miss emits `{ kind: 'absent' }` before returning `FS_NOT_FOUND`, while a file routes through `readText`/`streamText`, `buildWindow`, then emits `{ kind: 'present', version: info.version }`. The post-read confirming `stat` from the old `fileContext.read` stays dropped; a writer racing between the routing stat and the read can at worst make a later guarded edit spuriously stale.
|
||||
- **write** — `expectation = await ctx.waterfall('fs/write-intent', target, exec, () => undefined)`, then `ctx.fs.writeText(target, content, expectation)`, then emit the present outcome version. **Zero stat in the tool** with or without `dsh-fs-policy`.
|
||||
- **write** — `expectation = await ctx.waterfall('fs/write-intent', target, exec, () => undefined)`, then `ctx.fs.writeText(target, content, expectation)`, then emit the present outcome version. **Zero stat in the tool** with or without `dsh-fs-observation-policy`.
|
||||
- **edit** — `expectation = await ctx.waterfall('fs/edit-intent', target, exec, () => undefined)`, then `ctx.fs.editText(target, edit, expectation)`, then emit the present outcome version. **Zero stat in the tool** in both cases: the bare default is `undefined` (unconditional edit), so the tool never stats to manufacture a basis. If the target is absent on the bare path, the provider reports `FS_STALE_VERSION`; the policy returns `FS_NOT_FOUND` directly when it already holds an absent observation.
|
||||
|
||||
The tool passes `exec` (the tool-execution context) as the `actor` argument on every dispatch, so `dsh-fs-policy` can derive its observed-state owner. The tool does not know whether the policy plugin is present: it always provides the bare default behavior in the `next` thunk, and `dsh-fs-policy` short-circuits the thunk before it runs in the default deployment.
|
||||
The tool passes `exec` (the tool-execution context) as the `actor` argument on every dispatch, so `dsh-fs-observation-policy` can derive its observed-state owner. The tool does not know whether the policy plugin is present: it always provides the bare default behavior in the `next` thunk, and `dsh-fs-observation-policy` short-circuits the thunk before it runs in the default deployment.
|
||||
|
||||
**`fs/observed` fires after a successful operation and after a metadata probe confirms absence.** Its listeners must be synchronous, non-throwing recorders; the tool does not guard the plain emit, so a throwing listener can replace the pending read error or report failure after a mutation already succeeded. Async or fallible observation needs a separate event contract.
|
||||
|
||||
## Policy plugin contract (`dsh-fs-policy`)
|
||||
## Policy plugin contract (`dsh-fs-observation-policy`)
|
||||
|
||||
`dsh-fs-policy` is a plugin, not a service. It does not register `ctx.fileContext`, has no public method surface, and exposes no `read`/`write`/`edit`/`resolve` methods. It attaches three listeners via `ctx.on()` registrations (each returning a disposer for HMR). It keeps the observed-state `WeakMap<owner, Map<targetKey, FsObservation>>` and the structural owner derivation (narrowing the event's opaque `object` actor to its own `{ agent?: { session? } }` shape), but does not inject `fs` — every handler operates only on its own `WeakMap`, never on `ctx.fs`.
|
||||
`dsh-fs-observation-policy` is a plugin, not a service. It does not register `ctx.fileContext`, has no public method surface, and exposes no `read`/`write`/`edit`/`resolve` methods. It attaches three listeners via `ctx.on()` registrations (each returning a disposer for HMR). It keeps the observed-state `WeakMap<owner, Map<targetKey, FsObservation>>` and the structural owner derivation (narrowing the event's opaque `object` actor to its own `{ agent?: { session? } }` shape), but does not inject `fs` — every handler operates only on its own `WeakMap`, never on `ctx.fs`.
|
||||
|
||||
- `fs/write-intent` listener: unseen/absent ⇒ `createIfAbsent`; present ⇒ `replaceIfVersion`. It does NOT call `next()`: it fully owns the single decision slot.
|
||||
- `fs/edit-intent` listener: unseen ⇒ `FS_NOT_OBSERVED`; absent ⇒ `FS_NOT_FOUND`; present ⇒ its version guard. It does NOT call `next()`.
|
||||
@@ -138,17 +138,17 @@ The tool passes `exec` (the tool-execution context) as the `actor` argument on e
|
||||
|
||||
An observed-state entry is the **prior-observation record**, but its discriminant matters. Successful read/write/edit records present at a version, allowing create-then-edit or edit-then-edit without an intervening read. A read/view that confirms absence replaces any old positive version with absent, allowing only a guarded create; a later successful create replaces it with the new present version. Missing entry alone means unseen and produces `FS_NOT_OBSERVED` for edit. The owner is derived structurally from `{ agent?: { session? } }`; disposal drops all state (HMR safety).
|
||||
|
||||
`dsh-fs-policy` is now a pure policy/recording plugin with no service API — it influences the world only through the event gate. That is what removes the method coupling from `dsh-tool-fs`.
|
||||
`dsh-fs-observation-policy` is now a pure policy/recording plugin with no service API — it influences the world only through the event gate. That is what removes the method coupling from `dsh-tool-fs`.
|
||||
|
||||
## Bare-provider behavior (no `dsh-fs-policy`)
|
||||
## Bare-provider behavior (no `dsh-fs-observation-policy`)
|
||||
|
||||
This is not the intended deployment stance — a config loading the fs tools is expected to also load `dsh-fs-policy`. It is the unconstrained provider floor that exists once the tool is no longer coupled to a policy method service. With `dsh-fs-policy` absent, every `fs/*` waterfall falls through to its `undefined` default and `fs/observed` has no listener:
|
||||
This is not the intended deployment stance — a config loading the fs tools is expected to also load `dsh-fs-observation-policy`. It is the unconstrained provider floor that exists once the tool is no longer coupled to a policy method service. With `dsh-fs-observation-policy` absent, every `fs/*` waterfall falls through to its `undefined` default and `fs/observed` has no listener:
|
||||
|
||||
- **read** is identical (it never needed policy; it only emits a now-unheard `fs/observed`).
|
||||
- **write** unconditionally creates-or-overwrites: `expected` is `undefined`, so `writeText` writes whether or not the file exists and whatever its current version. No read-first requirement, no version check.
|
||||
- **edit** unconditionally replaces literal text in the file's current content: `expected` is `undefined`, so `editText` matches and rewrites without a version guard or a read-first requirement (`FS_EDIT_NOT_FOUND`/`FS_AMBIGUOUS_EDIT` still apply — those are about the literal match, not freshness). A missing target still reports `FS_STALE_VERSION`, matching the guarded edit path's "cannot edit this target now" code.
|
||||
|
||||
Both mutations are still atomic (the backend's per-target lock is unconditional). What is simply *absent*, not lost, is the policy `dsh-fs-policy` would add: observed-state, read-before-edit, and version-guarded write/edit. Loading `dsh-fs-policy` layers those constraints on by having its listeners return guarded `expected` values instead of `undefined`; nothing in the bare provider changes.
|
||||
Both mutations are still atomic (the backend's per-target lock is unconditional). What is simply *absent*, not lost, is the policy `dsh-fs-observation-policy` would add: observed-state, read-before-edit, and version-guarded write/edit. Loading `dsh-fs-observation-policy` layers those constraints on by having its listeners return guarded `expected` values instead of `undefined`; nothing in the bare provider changes.
|
||||
|
||||
## Supersedes
|
||||
|
||||
@@ -156,18 +156,18 @@ This amends — does not reverse — [the split-fs-seam Agent Note](../simplific
|
||||
|
||||
## Verification
|
||||
|
||||
Tests pin both paths: without `dsh-fs-policy`, the root tool plugin boots against `dsh-fs-local`, and read, create, overwrite, and unread edit succeed; with the policy, unread edit returns `FS_NOT_OBSERVED` and unread overwrite is gated by `createIfAbsent`. A later intent listener is not reached after the policy decides. Stale edits fail through provider CAS while the policy performs no `stat`; the tool budgets remain one `stat` for read and zero for write or edit on either path. The deletion recovery path is also assembled: stale mutation, missing reread, guarded recreation. Model-facing schemas remain byte-for-byte unchanged, while the recovered result transcript changes.
|
||||
Tests pin both paths: without `dsh-fs-observation-policy`, the root tool plugin boots against `dsh-fs-local`, and read, create, overwrite, and unread edit succeed; with the policy, unread edit returns `FS_NOT_OBSERVED` and unread overwrite is gated by `createIfAbsent`. A later intent listener is not reached after the policy decides. Stale edits fail through provider CAS while the policy performs no `stat`; the tool budgets remain one `stat` for read and zero for write or edit on either path. The deletion recovery path is also assembled: stale mutation, missing reread, guarded recreation. Model-facing schemas remain byte-for-byte unchanged, while the recovered result transcript changes.
|
||||
|
||||
## Alternatives considered
|
||||
|
||||
- **Keep `ctx.fileContext` as an in-path method service** — the shape [the split-fs-seam Agent Note](../simplification/2026-06-26-fsspec-style-fs-seam.md) first landed; rejected because the tool could not run without the policy layer, making policy load-bearing for basic operation instead of an opt-in tightening.
|
||||
- **Policy-side version checking** (`dsh-fs-policy` stats and compares in its waterfall handler) — rejected for the TOCTOU gap between that check and the tool's actual write; the provider's mutation critical section is the only race-free place, so the policy only chooses the CAS basis and gates on prior observation.
|
||||
- **Policy-side version checking** (`dsh-fs-observation-policy` stats and compares in its waterfall handler) — rejected for the TOCTOU gap between that check and the tool's actual write; the provider's mutation critical section is the only race-free place, so the policy only chooses the CAS basis and gates on prior observation.
|
||||
- **Per-tool `/read`/`/write`/`/edit` subpath plugins** — dropped on implementation: no consumer needed a single-tool deployment, and subpath publishing forced bespoke `tsdown`/`tsconfig`/`files`/workspace-constraint handling no sibling tool package carries; the per-tool registration helpers remain internal modules the root plugin composes.
|
||||
|
||||
## Consequences
|
||||
|
||||
- **Event indirection over a method call.** A waterfall + emit is less direct than `await ctx.fileContext.edit(...)`. The payoff is removing the tool-to-policy method dependency while keeping the default policy plugin; the cost is one more event vocabulary to learn. Mitigated by keeping the three events narrow and documenting the default-thunk semantics on each.
|
||||
- **Policy events in the storage seam.** `dsh-fs` gains two version-decision events plus a recording event though it is "just storage". This is the price of decoupling (the emitter cannot depend on the policy plugin). The events carry only `dsh-fs` vocabulary plus an opaque `object` actor and no model-facing concepts, so the seam stays free of line-window/observation policy types and of the agent/session owner structure.
|
||||
- **Single policy occupant, first-wins by convention.** The `fs/write-intent`/`fs/edit-intent` slots hold exactly one decider; the first-registered (or `prepend`ed) listener wins and the rest are short-circuited. `dsh-fs-policy` owning the slot is a deployment convention, not an event-enforced invariant — a second decider registered first would bypass it. This is acceptable because a second fs-version-policy decider is a misconfiguration, not a feature. If a future need for *layered* fs version policy appears, it is a new Agent Note (a composable value-passing waterfall), not a silent second listener on these events. Layered permission/audit/sandbox interception already has its home on `tools/execute`.
|
||||
- **Single policy occupant, first-wins by convention.** The `fs/write-intent`/`fs/edit-intent` slots hold exactly one decider; the first-registered (or `prepend`ed) listener wins and the rest are short-circuited. `dsh-fs-observation-policy` owning the slot is a deployment convention, not an event-enforced invariant — a second decider registered first would bypass it. This is acceptable because a second fs-version-policy decider is a misconfiguration, not a feature. If a future need for *layered* fs version policy appears, it is a new Agent Note (a composable value-passing waterfall), not a silent second listener on these events. Layered permission/audit/sandbox interception already has its home on `tools/execute`.
|
||||
- **Dropping the post-read confirming stat** makes a follow-up *guarded* edit occasionally fail-closed (`FS_STALE_VERSION` → re-read) under a read/write race. This is a UX nicety lost, never a correctness hole; the provider lock still prevents wrong-version writes.
|
||||
- **The bare provider does no read-before-write/edit and no version check.** A deployment without `dsh-fs-policy` lets the model overwrite or edit any existing file unconditionally. This is the deliberate meaning of keeping the tool independent of a policy service: the safety disciplines live in the `dsh-fs-policy` plugin. A deployment that omits it is opting into an unconstrained filesystem on purpose; that is not the intended stance for a config that ships the fs tools.
|
||||
- **The bare provider does no read-before-write/edit and no version check.** A deployment without `dsh-fs-observation-policy` lets the model overwrite or edit any existing file unconditionally. This is the deliberate meaning of keeping the tool independent of a policy service: the safety disciplines live in the `dsh-fs-observation-policy` plugin. A deployment that omits it is opting into an unconstrained filesystem on purpose; that is not the intended stance for a config that ships the fs tools.
|
||||
|
||||
@@ -1,4 +1,4 @@
|
||||
# Agent Note: 将 `dsh-fs-policy` 改为事件门禁插件，而非方法接口
|
||||
# Agent Note: 将 `dsh-fs-observation-policy` 改为事件门禁插件，而非方法接口
|
||||
|
||||
Status: implemented
|
||||
|
||||
@@ -11,19 +11,19 @@ Status: implemented
|
||||
这把三件本应可分离的事情耦合在了一起：
|
||||
|
||||
1. **工具做什么**——解析路径、读取窗口、写入/编辑文件。这是工具的职责，只需要 `ctx.fs`。
|
||||
2. **新鲜度/观测策略**——「编辑前必须先读」、「写入/编辑必须基于你读到的版本」。这是 `dsh-fs-policy` 插件的职责。
|
||||
2. **新鲜度/观测策略**——「编辑前必须先读」、「写入/编辑必须基于你读到的版本」。这是 `dsh-fs-observation-policy` 插件的职责。
|
||||
3. **观测状态的记录**——一个副作用，永远不应阻止工具正常运行。
|
||||
|
||||
由于工具调用的是 `fileContext` 方法，移除策略层就是一个破坏性变更，而非优雅地失去一个*附加*能力。策略层对工具的运行是承重性的，而非可选的收紧。
|
||||
|
||||
## 决策
|
||||
|
||||
反转控制流。**`dsh-tool-fs` 成为执行器，直接调用 `ctx.fs`**；**`dsh-fs-policy` 成为门控 + 记录插件**，通过事件参与，从不通过工具调用的方法，也不注册 `ctx.fileContext` 服务。
|
||||
反转控制流。**`dsh-tool-fs` 成为执行器，直接调用 `ctx.fs`**；**`dsh-fs-observation-policy` 成为门控 + 记录插件**，通过事件参与，从不通过工具调用的方法，也不注册 `ctx.fileContext` 服务。
|
||||
|
||||
```text
|
||||
tool dsh-tool-fs executor: resolves, reads windows, writes/edits via ctx.fs;
|
||||
emits fs policy events; renders results
|
||||
policy dsh-fs-policy plugin: listens to fs/write-intent +
|
||||
policy dsh-fs-observation-policy plugin: listens to fs/write-intent +
|
||||
fs/edit-intent (single-slot waterfall) and fs/observed
|
||||
(emit) events; adds observed-state + freshness.
|
||||
provider contract dsh-fs ctx.fs: text IO + ATOMIC mutation primitives whose version
|
||||
@@ -31,20 +31,20 @@ provider contract dsh-fs ctx.fs: text IO + ATOMIC mutation primitives
|
||||
provider dsh-fs-local local implementation of ctx.fs
|
||||
```
|
||||
|
||||
该模型是叠加式的：裸 `ctx.fs` 执行原子化、无约束的文本 I/O，而 `dsh-fs-policy` 叠加观测状态、先读后编辑和版本守卫。因此移除策略层后工具仍可用，只是不受约束。正式发布的 agent（智能体）配置会加载策略；裸模式的存在是为了让策略在服务边界保持可选，而非作为正常部署姿态。
|
||||
该模型是叠加式的：裸 `ctx.fs` 执行原子化、无约束的文本 I/O，而 `dsh-fs-observation-policy` 叠加观测状态、先读后编辑和版本守卫。因此移除策略层后工具仍可用，只是不受约束。正式发布的 agent（智能体）配置会加载策略；裸模式的存在是为了让策略在服务边界保持可选，而非作为正常部署姿态。
|
||||
|
||||
[文件系统缺失观测后续决策](../bug-fix/2026-08-09-filesystem-absence-observation.md)把记录载荷从仅表示成功的版本细化为显式的存在/缺失状态，并要求带防护的创建以不替换方式发布。事件门控归属与无 I/O 策略边界保持不变。
|
||||
|
||||
`dsh-tool-fs` 不再注入 `fileContext`。它注入 `fs` 和 `tools`/`systemPrompt`。
|
||||
|
||||
## 策略由提供方 CAS 强制执行，而非 `dsh-fs-policy` 的 stat
|
||||
## 策略由提供方 CAS 强制执行，而非 `dsh-fs-observation-policy` 的 stat
|
||||
|
||||
`dsh-fs-policy` 强制执行「你必须基于你读到的版本来写入/编辑」，**自身从不调用 `stat` 或比较版本**。它将观测到的版本作为 CAS 基准提供，让提供方的 mutation 临界区检测陈旧性：
|
||||
`dsh-fs-observation-policy` 强制执行「你必须基于你读到的版本来写入/编辑」，**自身从不调用 `stat` 或比较版本**。它将观测到的版本作为 CAS 基准提供，让提供方的 mutation 临界区检测陈旧性：
|
||||
|
||||
- 「该所有者最近观测到了什么？」是 `dsh-fs-policy` 在本地决定的唯一事项——一次 `WeakMap` 查找，无 I/O。无记录表示未见；缺失记录只允许带防护的创建；存在记录携带替换/编辑基准。
|
||||
- 「版本是否仍然有效，或者创建目标是否仍然缺失？」由**提供方的原子变更边界内部**决定。`dsh-fs-policy` 提供 `replaceIfVersion` 或 `createIfAbsent`；对于已经变化的版本，提供方抛出 `FS_STALE_VERSION`；带防护的创建若败给另一个创建者，则抛出 `FS_NOT_OBSERVED`。
|
||||
- 「该所有者最近观测到了什么？」是 `dsh-fs-observation-policy` 在本地决定的唯一事项——一次 `WeakMap` 查找，无 I/O。无记录表示未见；缺失记录只允许带防护的创建；存在记录携带替换/编辑基准。
|
||||
- 「版本是否仍然有效，或者创建目标是否仍然缺失？」由**提供方的原子变更边界内部**决定。`dsh-fs-observation-policy` 提供 `replaceIfVersion` 或 `createIfAbsent`；对于已经变化的版本，提供方抛出 `FS_STALE_VERSION`；带防护的创建若败给另一个创建者，则抛出 `FS_NOT_OBSERVED`。
|
||||
|
||||
这是有意为之的。如果 `dsh-fs-policy` 在其 waterfall（瀑布式事件）处理器中 stat 并比较版本，该检查与工具实际写入之间会存在 TOCTOU 间隙——文件可能在此期间变化，因此该检查只是一个虚假保证，提供方的锁无论如何都要兜底。将版本检查放在提供方的临界区中既无竞态又无额外 `stat`。所以 `dsh-fs-policy` **不做**任何文件系统 I/O；「必须基于最近一次读取」的保证由 CAS *实现*，`dsh-fs-policy` 只负责选择基准（`vObserved`）并对先前观测进行门控。
|
||||
这是有意为之的。如果 `dsh-fs-observation-policy` 在其 waterfall（瀑布式事件）处理器中 stat 并比较版本，该检查与工具实际写入之间会存在 TOCTOU 间隙——文件可能在此期间变化，因此该检查只是一个虚假保证，提供方的锁无论如何都要兜底。将版本检查放在提供方的临界区中既无竞态又无额外 `stat`。所以 `dsh-fs-observation-policy` **不做**任何文件系统 I/O；「必须基于最近一次读取」的保证由 CAS *实现*，`dsh-fs-observation-policy` 只负责选择基准（`vObserved`）并对先前观测进行门控。
|
||||
|
||||
## 提供方约定变更：版本守卫变为可选
|
||||
|
||||
@@ -54,7 +54,7 @@ provider dsh-fs-local local implementation of ctx.fs
|
||||
// writeText: expected is now optional. The FsWriteIntent union is UNCHANGED.
|
||||
writeText(target: FsTarget, content: string, expected?: FsWriteIntent, signal?: AbortSignal): Promise<FsWriteOutcome>
|
||||
// undefined → unconditionally create-or-overwrite (bare default)
|
||||
// createIfAbsent → create only, reject an existing file (dsh-fs-policy, unobserved) [unchanged]
|
||||
// createIfAbsent → create only, reject an existing file (dsh-fs-observation-policy, unobserved) [unchanged]
|
||||
// replaceIfVersion → overwrite only at the observed version, else FS_STALE_VERSION [unchanged]
|
||||
|
||||
// editText: expected becomes optional (was the required { version: FsVersion }).
|
||||
@@ -64,17 +64,17 @@ editText(target: FsTarget, edit: FsEditRequest, expected?: { version: FsVersion
|
||||
// { version } → edit only at that version, else FS_STALE_VERSION (the current behavior)
|
||||
```
|
||||
|
||||
`FsWriteIntent` 联合类型本身不变——第三种「无条件」状态通过*省略* `expected` 来表达，因此两个 mutation 共享同一种对称形状（`expected?`：省略 = 无守卫，传入 = 有守卫）。这对 `dsh-fs-policy` 使用的有守卫路径保持完全向后兼容；只有之前不可能出现的「无守卫」情况是新增的，且它是裸提供方的默认行为。无论哪种情况，mutation 仍在后端的 per-target 锁内运行，因此无条件写入/编辑仍是原子的（不会产生撕裂文件）；「无条件」去掉的是*版本*前置条件，而非原子性。`editText` 在有守卫和无守卫路径上都将缺失目标报告为 `FS_STALE_VERSION`，保持一个统一的编辑失败码表示「此刻无法编辑该目标」。
|
||||
`FsWriteIntent` 联合类型本身不变——第三种「无条件」状态通过*省略* `expected` 来表达，因此两个 mutation 共享同一种对称形状（`expected?`：省略 = 无守卫，传入 = 有守卫）。这对 `dsh-fs-observation-policy` 使用的有守卫路径保持完全向后兼容；只有之前不可能出现的「无守卫」情况是新增的，且它是裸提供方的默认行为。无论哪种情况，mutation 仍在后端的 per-target 锁内运行，因此无条件写入/编辑仍是原子的（不会产生撕裂文件）；「无条件」去掉的是*版本*前置条件，而非原子性。`editText` 在有守卫和无守卫路径上都将缺失目标报告为 `FS_STALE_VERSION`，保持一个统一的编辑失败码表示「此刻无法编辑该目标」。
|
||||
|
||||
## 事件词汇（由 `dsh-fs` 拥有）
|
||||
|
||||
事件定义在 `@deepseek-ai/dsh-fs` 中，而非 `dsh-fs-policy` 中。这是解耦约定所迫：`dsh-tool-fs` 是发射方，因此它必须引用事件类型，且即使 `dsh-fs-policy` 不再提供方法服务，它也必须能编译通过。`dsh-fs` 是 `dsh-tool-fs` 和 `dsh-fs-policy` 都已依赖的包，因此它是唯一能让发射方和策略监听方共享词汇而不让发射方依赖策略插件的归属地。
|
||||
事件定义在 `@deepseek-ai/dsh-fs` 中，而非 `dsh-fs-observation-policy` 中。这是解耦约定所迫：`dsh-tool-fs` 是发射方，因此它必须引用事件类型，且即使 `dsh-fs-observation-policy` 不再提供方法服务，它也必须能编译通过。`dsh-fs` 是 `dsh-tool-fs` 和 `dsh-fs-observation-policy` 都已依赖的包，因此它是唯一能让发射方和策略监听方共享词汇而不让发射方依赖策略插件的归属地。
|
||||
|
||||
这些事件携带既有的 `dsh-fs` 词汇（`FsTarget`、`FsVersion`、`FsObservation`、`FsWriteIntent`）加一个不透明的 actor——不携带面向模型的概念（行窗口、行号或渲染后的页脚不会泄漏到此层）。
|
||||
|
||||
**两个 `fs/*` 决策事件是单槽、先到先得的 waterfall。** `dsh-fs-policy` 不调用 `next()` 直接返回，因此在默认部署中它占据该槽位；更早注册或使用 `prepend` 的监听器会替代该策略。权限、审计和沙箱关注点仍留在可组合的 `tools/execute` waterfall 上。
|
||||
**两个 `fs/*` 决策事件是单槽、先到先得的 waterfall。** `dsh-fs-observation-policy` 不调用 `next()` 直接返回，因此在默认部署中它占据该槽位；更早注册或使用 `prepend` 的监听器会替代该策略。权限、审计和沙箱关注点仍留在可组合的 `tools/execute` waterfall 上。
|
||||
|
||||
actor 在 `dsh-fs` 中类型为 `object`——一个纯粹的不透明载体，提供方约定从不读取或收窄它。owner 的推导（`actor.agent?.session`）和 `{ agent?: { session? } }` 结构形状完全留在 `dsh-fs-policy` 内部，由其在监听器中将 `object` actor 收窄为该形状。`dsh-fs` 拥有事件名和 fs 词汇；它不拥有策略层的运行时 owner 结构。
|
||||
actor 在 `dsh-fs` 中类型为 `object`——一个纯粹的不透明载体，提供方约定从不读取或收窄它。owner 的推导（`actor.agent?.session`）和 `{ agent?: { session? } }` 结构形状完全留在 `dsh-fs-observation-policy` 内部，由其在监听器中将 `object` actor 收窄为该形状。`dsh-fs` 拥有事件名和 fs 词汇；它不拥有策略层的运行时 owner 结构。
|
||||
|
||||
```ts
|
||||
import type { FsObservation, FsTarget, FsVersion, FsWriteIntent } from '@deepseek-ai/dsh-fs'
|
||||
@@ -99,7 +99,7 @@ interface Events {
|
||||
/**
|
||||
* Record that an actor observed a target as present at a version or absent.
|
||||
* Fire-and-forget (plain emit). Listeners MUST be
|
||||
* synchronous, side-effect-only recorders (`dsh-fs-policy`'s is a WeakMap
|
||||
* synchronous, side-effect-only recorders (`dsh-fs-observation-policy`'s is a WeakMap
|
||||
* write); the tool does not guard the emit, so a throwing listener surfaces as
|
||||
* the tool's isError result. No listener ⇒ nothing recorded.
|
||||
* @mode emit
|
||||
@@ -112,7 +112,7 @@ interface Events {
|
||||
|
||||
## 工具约定（`dsh-tool-fs`）
|
||||
|
||||
工具保留其面向模型的 schema（`read`/`write`/`edit`，逐字节不变）和提示词段落。提示词引导仍以策略优先，因为加载 fs 工具的部署预期也会加载 `dsh-fs-policy`：模型仍被告知在覆写或编辑前先读取，而该要求来自 fs-policy 插件，并非后端。裸提供方回退不改变提示词立场。
|
||||
工具保留其面向模型的 schema（`read`/`write`/`edit`，逐字节不变）和提示词段落。提示词引导仍以策略优先，因为加载 fs 工具的部署预期也会加载 `dsh-fs-observation-policy`：模型仍被告知在覆写或编辑前先读取，而该要求来自 fs-observation-policy 插件，并非后端。裸提供方回退不改变提示词立场。
|
||||
|
||||
`dsh-tool-fs` 获得从旧 `fileContext` 方法服务迁移来的执行器职责，包括**读取渲染**（`read-render.ts`：`buildWindow` + `formatReadOutput`、`READ_MAX_BYTES`、`READ_MAX_LINE_LENGTH`、`FileReadOutcome`/`FileTextLine`，以及 `read.ts` 中的 `STREAM_MIN_SIZE`），这些现在是工具的渲染细节，因为读取已由工具拥有。这些读取渲染类型和辅助函数移入 `dsh-tool-fs`；策略插件不得继续作为工具的类型依赖。
|
||||
|
||||
@@ -121,16 +121,16 @@ interface Events {
|
||||
通过让 waterfall 惰性产出期望值来最小化 `stat` 预算——裸默认返回 `undefined`（无守卫），从不 stat：
|
||||
|
||||
- **read**——一次 `stat`；元数据未命中时，在返回 `FS_NOT_FOUND` 前 emit `{ kind: 'absent' }`；目标为文件时，则依次执行 `readText`/`streamText`、`buildWindow`，再 emit `{ kind: 'present', version: info.version }`。旧 `fileContext.read` 中读后确认的 `stat` 仍保持移除；在路由 stat 和读取之间竞争的写入者最多只能使后续带防护的编辑误报陈旧。
|
||||
- **write**——`expectation = await ctx.waterfall('fs/write-intent', target, exec, () => undefined)`，然后 `ctx.fs.writeText(target, content, expectation)`，再 emit 表示存在的结果版本。无论是否有 `dsh-fs-policy`，**工具内零 stat**。
|
||||
- **write**——`expectation = await ctx.waterfall('fs/write-intent', target, exec, () => undefined)`，然后 `ctx.fs.writeText(target, content, expectation)`，再 emit 表示存在的结果版本。无论是否有 `dsh-fs-observation-policy`，**工具内零 stat**。
|
||||
- **edit**——`expectation = await ctx.waterfall('fs/edit-intent', target, exec, () => undefined)`，然后 `ctx.fs.editText(target, edit, expectation)`，再 emit 表示存在的结果版本。两种情况下**工具内零 stat**：裸默认为 `undefined`（无条件编辑），因此工具从不 stat 来制造基准。如果裸路径上的目标不存在，提供方报告 `FS_STALE_VERSION`；策略已持有缺失观测时，则直接返回 `FS_NOT_FOUND`。
|
||||
|
||||
工具在每次分发时将 `exec`（工具执行上下文）作为 `actor` 参数传入，以便 `dsh-fs-policy` 推导其观测状态的 owner。工具不知道策略插件是否存在：它始终在 `next` thunk 中提供裸默认行为，而 `dsh-fs-policy` 在默认部署中会在 thunk 运行前短路它。
|
||||
工具在每次分发时将 `exec`（工具执行上下文）作为 `actor` 参数传入，以便 `dsh-fs-observation-policy` 推导其观测状态的 owner。工具不知道策略插件是否存在：它始终在 `next` thunk 中提供裸默认行为，而 `dsh-fs-observation-policy` 在默认部署中会在 thunk 运行前短路它。
|
||||
|
||||
**`fs/observed` 在操作成功后，以及元数据探测确认缺失后触发。** 其监听器必须是同步、不抛异常的记录器；工具不对 plain emit 做保护，因此抛异常的监听器可能取代待返回的读取错误，或在 mutation 已成功后报告失败。异步或可失败的观测需要另一份事件约定。
|
||||
|
||||
## 策略插件约定（`dsh-fs-policy`）
|
||||
## 策略插件约定（`dsh-fs-observation-policy`）
|
||||
|
||||
`dsh-fs-policy` 是插件，不是服务。它不注册 `ctx.fileContext`，没有公开方法面，不暴露 `read`/`write`/`edit`/`resolve` 方法。它通过 `ctx.on()` 注册三个监听器（每个返回一个 disposer 用于 HMR（热模块替换））。它维护观测状态 `WeakMap<owner, Map<targetKey, FsObservation>>`，以及结构化的 owner 推导（将事件中不透明的 `object` actor 收窄为自己的 `{ agent?: { session? } }` 形状），但不注入 `fs`——每个处理器只操作自己的 `WeakMap`，从不操作 `ctx.fs`。
|
||||
`dsh-fs-observation-policy` 是插件，不是服务。它不注册 `ctx.fileContext`，没有公开方法面，不暴露 `read`/`write`/`edit`/`resolve` 方法。它通过 `ctx.on()` 注册三个监听器（每个返回一个 disposer 用于 HMR（热模块替换））。它维护观测状态 `WeakMap<owner, Map<targetKey, FsObservation>>`，以及结构化的 owner 推导（将事件中不透明的 `object` actor 收窄为自己的 `{ agent?: { session? } }` 形状），但不注入 `fs`——每个处理器只操作自己的 `WeakMap`，从不操作 `ctx.fs`。
|
||||
|
||||
- `fs/write-intent` 监听器：未见/缺失 ⇒ `createIfAbsent`；存在 ⇒ `replaceIfVersion`。它不调用 `next()`：完全占据单一决策槽位。
|
||||
- `fs/edit-intent` 监听器：未见 ⇒ `FS_NOT_OBSERVED`；缺失 ⇒ `FS_NOT_FOUND`；存在 ⇒ 返回其版本守卫。同样不调用 `next()`。
|
||||
@@ -138,17 +138,17 @@ interface Events {
|
||||
|
||||
一条观测状态条目是**先前观测记录**，但其可辨识字段会影响决策。成功的 read/write/edit 会记录存在状态及版本，使 create-then-edit 或 edit-then-edit 序列无需中间重新读取即可工作。确认缺失的 read/view 会用缺失状态取代旧的正向版本，因此只允许带防护的创建；随后成功的创建会再用新的存在版本取代缺失状态。只有条目不存在才表示未见，并使 edit 返回 `FS_NOT_OBSERVED`。owner 从 `{ agent?: { session? } }` 结构化推导；dispose 时丢弃所有状态（HMR 安全）。
|
||||
|
||||
`dsh-fs-policy` 现在是一个纯策略/记录插件，没有服务 API——它只通过事件门禁影响外界。这正是移除 `dsh-tool-fs` 方法耦合的关键。
|
||||
`dsh-fs-observation-policy` 现在是一个纯策略/记录插件，没有服务 API——它只通过事件门禁影响外界。这正是移除 `dsh-tool-fs` 方法耦合的关键。
|
||||
|
||||
## 裸提供方行为（无 `dsh-fs-policy`）
|
||||
## 裸提供方行为（无 `dsh-fs-observation-policy`）
|
||||
|
||||
这不是预期的部署姿态——加载 fs 工具的配置预期也会加载 `dsh-fs-policy`。它是工具不再耦合于策略方法服务后所存在的无约束提供方下限。当 `dsh-fs-policy` 不存在时，每个 `fs/*` waterfall 落入其 `undefined` 默认值，`fs/observed` 无监听器：
|
||||
这不是预期的部署姿态——加载 fs 工具的配置预期也会加载 `dsh-fs-observation-policy`。它是工具不再耦合于策略方法服务后所存在的无约束提供方下限。当 `dsh-fs-observation-policy` 不存在时，每个 `fs/*` waterfall 落入其 `undefined` 默认值，`fs/observed` 无监听器：
|
||||
|
||||
- **read** 行为不变（它从不需要策略；只是 emit 了一个现在无人监听的 `fs/observed`）。
|
||||
- **write** 无条件 create-or-overwrite：`expected` 为 `undefined`，因此 `writeText` 无论文件是否存在、无论当前版本如何都直接写入。无先读要求，无版本检查。
|
||||
- **edit** 无条件替换文件当前内容中的字面文本：`expected` 为 `undefined`，因此 `editText` 无版本守卫、无先读要求地匹配并重写（`FS_EDIT_NOT_FOUND`/`FS_AMBIGUOUS_EDIT` 仍适用——它们关乎字面匹配，而非新鲜度）。缺失目标仍报告 `FS_STALE_VERSION`，与有守卫编辑路径的「此刻无法编辑该目标」错误码一致。
|
||||
|
||||
两个 mutation 仍是原子的（后端的 per-target 锁是无条件的）。仅仅是*不存在*（而非丢失）的是 `dsh-fs-policy` 本会叠加的策略：观测状态、先读后编辑和版本守卫的写入/编辑。加载 `dsh-fs-policy` 后，其监听器返回有守卫的 `expected` 值而非 `undefined`，从而叠加这些约束；裸提供方本身无需任何变更。
|
||||
两个 mutation 仍是原子的（后端的 per-target 锁是无条件的）。仅仅是*不存在*（而非丢失）的是 `dsh-fs-observation-policy` 本会叠加的策略：观测状态、先读后编辑和版本守卫的写入/编辑。加载 `dsh-fs-observation-policy` 后，其监听器返回有守卫的 `expected` 值而非 `undefined`，从而叠加这些约束；裸提供方本身无需任何变更。
|
||||
|
||||
## 取代关系
|
||||
|
||||
@@ -156,18 +156,18 @@ interface Events {
|
||||
|
||||
## 验证
|
||||
|
||||
测试固定了两条路径：无 `dsh-fs-policy` 时，根工具插件对 `dsh-fs-local` 启动，read、create、overwrite 和未读 edit 均成功；有策略时，未读 edit 返回 `FS_NOT_OBSERVED`，未读 overwrite 被 `createIfAbsent` 门控。策略决定后，后注册的 intent 监听器不会被触达。陈旧编辑通过提供方 CAS 失败，而策略不执行 `stat`；工具预算在两条路径上保持 read 一次 `stat`，write 或 edit 均为零次。测试也组装了删除恢复路径：陈旧变更、重新读取时确认缺失、带防护的重新创建。面向模型的 schema 逐字节不变，但恢复后的结果 transcript（文本记录）发生变化。
|
||||
测试固定了两条路径：无 `dsh-fs-observation-policy` 时，根工具插件对 `dsh-fs-local` 启动，read、create、overwrite 和未读 edit 均成功；有策略时，未读 edit 返回 `FS_NOT_OBSERVED`，未读 overwrite 被 `createIfAbsent` 门控。策略决定后，后注册的 intent 监听器不会被触达。陈旧编辑通过提供方 CAS 失败，而策略不执行 `stat`；工具预算在两条路径上保持 read 一次 `stat`，write 或 edit 均为零次。测试也组装了删除恢复路径：陈旧变更、重新读取时确认缺失、带防护的重新创建。面向模型的 schema 逐字节不变，但恢复后的结果 transcript（文本记录）发生变化。
|
||||
|
||||
## 曾考虑的替代方案
|
||||
|
||||
- **保留 `ctx.fileContext` 作为关键路径上的方法服务**——[拆分文件系统 seam Agent Note](../simplification/2026-06-26-fsspec-style-fs-seam.md) 最初落地的形态；否决，因为工具无法在没有策略层的情况下运行，使策略对基本操作是承重性的，而非可选的收紧。
|
||||
- **策略侧版本检查**（`dsh-fs-policy` 在其 waterfall 处理器中 stat 并比较版本）——否决，因为该检查与工具实际写入之间存在 TOCTOU 间隙；提供方的 mutation 临界区是唯一无竞态的位置，因此策略只选择 CAS 基准并对先前观测进行门控。
|
||||
- **策略侧版本检查**（`dsh-fs-observation-policy` 在其 waterfall 处理器中 stat 并比较版本）——否决，因为该检查与工具实际写入之间存在 TOCTOU 间隙；提供方的 mutation 临界区是唯一无竞态的位置，因此策略只选择 CAS 基准并对先前观测进行门控。
|
||||
- **每工具 `/read`/`/write`/`/edit` 子路径插件**——实现时放弃：没有消费方需要单工具部署，且子路径发布迫使引入兄弟工具包都不需要的定制 `tsdown`/`tsconfig`/`files`/workspace-constraint 处理；每工具的注册辅助函数仍作为根插件组合的内部模块保留。
|
||||
|
||||
## 后果
|
||||
|
||||
- **事件间接层取代方法调用。** 一次 waterfall + emit 不如 `await ctx.fileContext.edit(...)` 直接。收益是移除了工具到策略的方法依赖，同时保留默认策略插件；代价是多一套事件词汇需要学习。通过保持三个事件的窄小范围并在每个事件上记录 default-thunk 语义来缓解。
|
||||
- **策略事件位于存储 seam 中。** `dsh-fs` 增加了两个版本决策事件和一个记录事件，尽管它「只是存储」。这是解耦的代价（发射方不能依赖策略插件）。这些事件只携带 `dsh-fs` 词汇加一个不透明的 `object` actor，不携带面向模型的概念，因此 seam 不沾染行窗口/观测策略类型，也不沾染 agent/会话所有者结构。
|
||||
- **单一策略占位者，按惯例先到先得。** `fs/write-intent`/`fs/edit-intent` 槽位恰好容纳一个决策者；先注册（或 `prepend`）的监听器获胜，其余被短路。`dsh-fs-policy` 占据该槽位是部署惯例，而非事件系统强制的不变式——一个先注册的第二决策者会绕过它。这是可接受的，因为第二个 fs 版本策略决策者是配置错误，而非功能。如果未来出现*分层* fs 版本策略的需求，那是一个新 Agent Note（可组合的值传递 waterfall），而非在这些事件上静默添加第二个监听器。分层的权限/审计/沙箱拦截已有其归属：`tools/execute`。
|
||||
- **单一策略占位者，按惯例先到先得。** `fs/write-intent`/`fs/edit-intent` 槽位恰好容纳一个决策者；先注册（或 `prepend`）的监听器获胜，其余被短路。`dsh-fs-observation-policy` 占据该槽位是部署惯例，而非事件系统强制的不变式——一个先注册的第二决策者会绕过它。这是可接受的，因为第二个 fs 版本策略决策者是配置错误，而非功能。如果未来出现*分层* fs 版本策略的需求，那是一个新 Agent Note（可组合的值传递 waterfall），而非在这些事件上静默添加第二个监听器。分层的权限/审计/沙箱拦截已有其归属：`tools/execute`。
|
||||
- **移除读后确认 stat** 使后续*有守卫*的编辑在 read/write 竞争下偶尔为安全起见拒绝写入（`FS_STALE_VERSION` → 重新读取）。这是丢失的 UX 便利，绝非正确性漏洞；提供方锁仍阻止基于错误版本的写入。
|
||||
- **裸提供方不做先读后写/编辑，也不做版本检查。** 没有 `dsh-fs-policy` 的部署允许模型无条件覆写或编辑任何已有文件。这正是保持工具独立于策略服务的有意含义：安全纪律存在于 `dsh-fs-policy` 插件中。省略它的部署是有意选择无约束的文件系统；对于发布 fs 工具的配置而言，这不是预期的姿态。
|
||||
- **裸提供方不做先读后写/编辑，也不做版本检查。** 没有 `dsh-fs-observation-policy` 的部署允许模型无条件覆写或编辑任何已有文件。这正是保持工具独立于策略服务的有意含义：安全纪律存在于 `dsh-fs-observation-policy` 插件中。省略它的部署是有意选择无约束的文件系统；对于发布 fs 工具的配置而言，这不是预期的姿态。
|
||||
|
||||
@@ -2,5 +2,5 @@
|
||||
# side as of the last confirmed-consistent state. Both languages carry equal authority;
|
||||
# after editing either side, bring the other along and re-record with:
|
||||
# pnpm run verify-translation-pairing --write .agents/notes/implemented/architecture/2026-06-30-bash-stdin-env-trusted-plugin-api.md
|
||||
2026-06-30-bash-stdin-env-trusted-plugin-api.md: 2087b2f7a9682ad5f554d4a7a2f521485d164432
|
||||
2026-06-30-bash-stdin-env-trusted-plugin-api.zh.md: e5a4303d64b9ef1445a8b9b380a0ed83339c1d48
|
||||
2026-06-30-bash-stdin-env-trusted-plugin-api.md: 41be63fff598587ee9b873cf7edf51da788bc02a
|
||||
2026-06-30-bash-stdin-env-trusted-plugin-api.zh.md: 4f0338cbb6a265ff629a81e9c9a12c4456941ba0
|
||||
|
||||
@@ -6,13 +6,13 @@ English | [中文](2026-06-30-bash-stdin-env-trusted-plugin-api.zh.md)
|
||||
|
||||
## Problem
|
||||
|
||||
The hooks subsystem runs external hook commands the way Claude Code and Codex do: a hook is a shell command that receives its event payload as **JSON on stdin** and reads context from a handful of **environment variables** (`CLAUDE_PROJECT_DIR`, `CLAUDE_PLUGIN_ROOT`, `PLUGIN_ROOT`, …). The harness already has a perfectly good command runner behind the `ctx.bash` capability seam ([dsh-bash](../../../../packages/bash/bash) → [dsh-bash-local](../../../../packages/bash/bash-local)), with process-group kills, output truncation/spill, and a credential scrub. Reusing it for hook execution means a hook bridge does not re-implement subprocess plumbing — but the seam had no way to write stdin or set extra env. This change adds those two inputs.
|
||||
The hooks subsystem runs external hook commands the way Claude Code and Codex do: a hook is a shell command that receives its event payload as **JSON on stdin** and reads context from a handful of **environment variables** (`CLAUDE_PROJECT_DIR`, `CLAUDE_PLUGIN_ROOT`, `PLUGIN_ROOT`, …). The harness already has a perfectly good command runner behind the `ctx.shell` capability seam ([dsh-shell](../../../../packages/shell/shell) → [dsh-bash-local](../../../../packages/shell/bash-local)), with process-group kills, output truncation/spill, and a credential scrub. Reusing it for hook execution means a hook bridge does not re-implement subprocess plumbing — but the seam had no way to write stdin or set extra env. This change adds those two inputs.
|
||||
|
||||
`stdin` and `env` do not create a new model capability because ordinary shell syntax already supplies both. Ambient credentials are protected by `dsh-bash-local`'s child-environment scrub, not by hiding these Service Definition fields; model tool arguments are static JSON and do not expand shell variables. The fields therefore serve trusted in-process callers, such as hook bridges, that need to pass structured input and `CLAUDE_*` variables without embedding them in model-visible shell text. See [defensive-patterns.md](../../../../docs/defensive-patterns.md) for the ambient-environment rule.
|
||||
|
||||
## Decision
|
||||
|
||||
Add `stdin?: string` and `env?: Record<string, string>` to **both** `BashExecRequest` (the model-/plugin-facing request) and `BashExecSpec` (the resolved spec `run`/`start` act on), and thread them through `dsh-bash-local`: `resolve()` carries them verbatim, `run()`/`start()` pass them to `runBash`, which writes the bytes to the child's stdin and merges the extra env.
|
||||
Add `stdin?: string` and `env?: Record<string, string>` to **both** `ShellExecRequest` (the model-/plugin-facing request) and `ShellExecSpec` (the resolved spec `run`/`start` act on), and thread them through `dsh-bash-local`: `resolve()` carries them verbatim, `run()`/`start()` pass them to `runBash`, which writes the bytes to the child's stdin and merges the extra env.
|
||||
|
||||
Three deliberate choices:
|
||||
|
||||
@@ -30,4 +30,4 @@ Three deliberate choices:
|
||||
|
||||
## Consequences
|
||||
|
||||
Hook bridges pass JSON payloads and hook-specific variables through the existing bash seam, retaining its process-group, truncation, and spill behavior. The model-facing behavior remains unchanged, and the bash tool remains the sole owner of model-call request construction. The vocabulary lives in [the bash data-structure reference](../../../../docs/subsystems/bash.md).
|
||||
Hook bridges pass JSON payloads and hook-specific variables through the existing bash seam, retaining its process-group, truncation, and spill behavior. The model-facing behavior remains unchanged, and the bash tool remains the sole owner of model-call request construction. The vocabulary lives in [the bash data-structure reference](../../../../docs/subsystems/shell.md).
|
||||
|
||||
@@ -6,13 +6,13 @@ Status: implemented
|
||||
|
||||
## 问题
|
||||
|
||||
钩子子系统以 Claude Code 和 Codex 的方式运行外部钩子命令：钩子是一条 shell 命令，通过 **stdin 上的 JSON** 接收事件载荷，并从若干**环境变量**（`CLAUDE_PROJECT_DIR`、`CLAUDE_PLUGIN_ROOT`、`PLUGIN_ROOT`……）读取上下文。harness 已经在 `ctx.bash` 能力 seam 后面有一个完善的命令执行器（[dsh-bash](../../../../packages/bash/bash) → [dsh-bash-local](../../../../packages/bash/bash-local)），具备进程组终止、输出截断/spill 处理和凭证擦除功能。复用它来执行钩子意味着钩子桥接层无需重新实现子进程底层机制——但该 seam 此前无法写入 stdin 或设置额外 env。本次变更添加这两个输入。
|
||||
钩子子系统以 Claude Code 和 Codex 的方式运行外部钩子命令：钩子是一条 shell 命令，通过 **stdin 上的 JSON** 接收事件载荷，并从若干**环境变量**（`CLAUDE_PROJECT_DIR`、`CLAUDE_PLUGIN_ROOT`、`PLUGIN_ROOT`……）读取上下文。harness 已经在 `ctx.shell` 能力 seam 后面有一个完善的命令执行器（[dsh-shell](../../../../packages/shell/shell) → [dsh-bash-local](../../../../packages/shell/bash-local)），具备进程组终止、输出截断/spill 处理和凭证擦除功能。复用它来执行钩子意味着钩子桥接层无需重新实现子进程底层机制——但该 seam 此前无法写入 stdin 或设置额外 env。本次变更添加这两个输入。
|
||||
|
||||
`stdin` 和 `env` 不构成新的模型能力，因为普通 shell 语法已经能提供两者。环境凭证由 `dsh-bash-local` 的子环境擦除机制保护，而非靠隐藏这些 Service Definition 字段；模型工具参数是静态 JSON，不会展开 shell 变量。因此这些字段服务于受信的进程内调用方（如钩子桥接层），它们需要传递结构化输入和 `CLAUDE_*` 变量，而不必将其嵌入模型可见的 shell 文本。环境变量规则见 [defensive-patterns.md](../../../../docs/defensive-patterns.md)。
|
||||
|
||||
## 决策
|
||||
|
||||
在 `BashExecRequest`（模型/插件侧请求）和 `BashExecSpec`（`run`/`start` 所作用的已解析 spec）上**同时**添加 `stdin?: string` 与 `env?: Record<string, string>`，并在 `dsh-bash-local` 中贯穿它们：`resolve()` 原样传递，`run()`/`start()` 将其传给 `runBash`，后者把字节写入子进程的 stdin 并合并额外 env。
|
||||
在 `ShellExecRequest`（模型/插件侧请求）和 `ShellExecSpec`（`run`/`start` 所作用的已解析 spec）上**同时**添加 `stdin?: string` 与 `env?: Record<string, string>`，并在 `dsh-bash-local` 中贯穿它们：`resolve()` 原样传递，`run()`/`start()` 将其传给 `runBash`，后者把字节写入子进程的 stdin 并合并额外 env。
|
||||
|
||||
三个有意为之的选择：
|
||||
|
||||
@@ -30,4 +30,4 @@ Status: implemented
|
||||
|
||||
## 后果
|
||||
|
||||
钩子桥接层通过既有的 bash seam 传递 JSON 载荷和钩子特定变量，保留其进程组终止、截断和 spill 行为。面向模型的行为不变，bash 工具仍是模型调用请求构建的唯一所有者。相关词汇定义见 [bash 数据结构参考](../../../../docs/subsystems/bash.md)。
|
||||
钩子桥接层通过既有的 bash seam 传递 JSON 载荷和钩子特定变量，保留其进程组终止、截断和 spill 行为。面向模型的行为不变，bash 工具仍是模型调用请求构建的唯一所有者。相关词汇定义见 [bash 数据结构参考](../../../../docs/subsystems/shell.md)。
|
||||
|
||||
@@ -2,5 +2,5 @@
|
||||
# side as of the last confirmed-consistent state. Both languages carry equal authority;
|
||||
# after editing either side, bring the other along and re-record with:
|
||||
# pnpm run verify-translation-pairing --write .agents/notes/implemented/architecture/2026-07-02-tool-render-intent-union.md
|
||||
2026-07-02-tool-render-intent-union.md: 67607b2848305439513503d7e03ad5e2a2e4020a
|
||||
2026-07-02-tool-render-intent-union.zh.md: f69442cc1cf226a5dd949fc3890029a69a639953
|
||||
2026-07-02-tool-render-intent-union.md: 52626d3d9200146df95aee7836d37bbaf7e6a9ec
|
||||
2026-07-02-tool-render-intent-union.zh.md: 43fda366694aecda07b276d34c882e75d6b4aa95
|
||||
|
||||
@@ -47,14 +47,14 @@ interface TerminalResultView { card: 'terminal'; title?: string; output?: string
|
||||
### Producer mapping
|
||||
|
||||
- `dsh-tool-fs` read → `generic` (`kind:'read'`, a follow-along `location`); write → `diff` (`oldText:null`); edit → `diff` (`oldText:old_string || null`, `newText:new_string ?? ''`). This mirrors `claude-agent-acp`'s `toolInfoFromToolUse` Read/Write/Edit arms field-for-field.
|
||||
- `dsh-tool-bash` foreground → `terminal` call + `terminal` result; `run_in_background` → `generic`. The generic `task_*` controls own their own generic cards.
|
||||
- `dsh-tool-bash` foreground → `terminal` call + `terminal` result; `run_in_background` → `generic`. The generic `job_*` controls own their own generic cards.
|
||||
- `dsh-tool-todo` → `generic`.
|
||||
|
||||
### Terminal fallback ownership
|
||||
|
||||
`TerminalResultView` carries only `output`/`exitCode`/`signal`. A UI without the terminal capability needs a fenced ` ```console ` text fallback; that derivation moves to the **bridge** (it wraps `output` in a fenced block on the no-capability path), rather than the tool double-encoding it. This keeps the bash tool's result a single structured shape and preserves the existing capability-gated behavior byte-for-byte.
|
||||
|
||||
The terminal intent is display-only. The harness still executes the command through its bash service, preserving sandboxing, environment scrubbing, task ownership, and per-session cwd; a UI projects the completed call and never becomes a second execution backend.
|
||||
The terminal intent is display-only. The harness still executes the command through its bash service, preserving sandboxing, environment scrubbing, job ownership, and per-session cwd; a UI projects the completed call and never becomes a second execution backend.
|
||||
|
||||
### Purity preserved
|
||||
|
||||
|
||||
@@ -47,7 +47,7 @@ interface TerminalResultView { card: 'terminal'; title?: string; output?: string
|
||||
### 生产者映射
|
||||
|
||||
- `dsh-tool-fs` read → `generic`（`kind:'read'`，附带一个 follow-along `location`）；write → `diff`（`oldText:null`）；edit → `diff`（`oldText:old_string || null`，`newText:new_string ?? ''`）。这与 `claude-agent-acp` 的 `toolInfoFromToolUse` 中 Read/Write/Edit 各分支逐字段对应。
|
||||
- `dsh-tool-bash` 前台运行 → `terminal` 调用 + `terminal` 结果；`run_in_background` → `generic`。通用 `task_*` 控制工具拥有各自的 generic 卡片。
|
||||
- `dsh-tool-bash` 前台运行 → `terminal` 调用 + `terminal` 结果；`run_in_background` → `generic`。通用 `job_*` 控制工具拥有各自的 generic 卡片。
|
||||
- `dsh-tool-todo` → `generic`。
|
||||
|
||||
### 终端回退的归属
|
||||
|
||||
@@ -2,5 +2,5 @@
|
||||
# side as of the last confirmed-consistent state. Both languages carry equal authority;
|
||||
# after editing either side, bring the other along and re-record with:
|
||||
# pnpm run verify-translation-pairing --write .agents/notes/implemented/architecture/2026-07-05-prompt-variables-and-tool-guidance-ownership.md
|
||||
2026-07-05-prompt-variables-and-tool-guidance-ownership.md: 8f8b40a44314fc6add97842d48273e49927f92e2
|
||||
2026-07-05-prompt-variables-and-tool-guidance-ownership.zh.md: 53f8b4b6fbfd0e3acb9ad803db19c609d2d718b2
|
||||
2026-07-05-prompt-variables-and-tool-guidance-ownership.md: 9a53619d9510e3f4fa561f8420b2da3bedbbf4bb
|
||||
2026-07-05-prompt-variables-and-tool-guidance-ownership.zh.md: 6174f277f0a4e8c46ff8d732411f28b4ea97728c
|
||||
|
||||
@@ -10,7 +10,7 @@ The assembled system prompt had four defects, all of one family: facts the harne
|
||||
|
||||
**The model could not know its own name.** `AgentOptions.model` drives every request, but no prompt text carried it — and nothing COULD carry it: sections in `dsh-system-prompt` were context-global while the model name is per-agent, and `assemble()` took no per-agent input at all.
|
||||
|
||||
**Tool guidance was hand-written prose in leaf YAML.** The bash/subagent/todo_write usage guidance lived in the coding-agent and ACP persona strings — two drifting copies (the ACP one was already abridged) — while `dsh-tool-fs` and `dsh-tool-web` owned their guidance as `ctx.systemPrompt.section()` contributions. Loading or dropping a tool plugin meant editing every deployment's persona by hand, and the old terminal welcome banner hand-enumerated the tool set too.
|
||||
**Tool guidance was hand-written prose in leaf YAML.** The shell/subagent/todo_write usage guidance lived in the coding-agent and ACP persona strings — two drifting copies (the ACP one was already abridged) — while `dsh-tool-fs` and `dsh-tool-web` owned their guidance as `ctx.systemPrompt.section()` contributions. Loading or dropping a tool plugin meant editing every deployment's persona by hand, and the old terminal welcome banner hand-enumerated the tool set too.
|
||||
|
||||
**The persona rendered after tool guidance.** The loop string-joined `agent.options.systemPrompt` AFTER the assembled sections, so the model read "Use the read tool…" before "You are a coding agent" — backwards relative to the identity-first convention (Claude Code, Codex) and a second composition path besides the section pipeline.
|
||||
|
||||
@@ -58,7 +58,7 @@ Per-tool semantics and selection guidance live in tool descriptions. Prompt sect
|
||||
|
||||
## Shipped invariants
|
||||
|
||||
- The tui-agent prompt renders identity, persona with the interpolated model, then fs/bash/web guidance through one assembly path.
|
||||
- The tui-agent prompt renders identity, persona with the interpolated model, then fs/shell/web guidance through one assembly path.
|
||||
- Fork and fresh subagent descriptions reflect whether the provider inherits completed conversation turns; the tool appears, disappears, and is reworded with provider lifecycle changes.
|
||||
- Unknown, valueless, malformed, or unbalanced variable references name the section and throw; duplicate section, variable, and tool registrations also throw.
|
||||
- Snapshot replay is prompt-independent: it keys recorded chunk streams by turn and step without comparing the outgoing request.
|
||||
|
||||
@@ -10,7 +10,7 @@ Status: implemented
|
||||
|
||||
**模型无法知道自己的名字。** `AgentOptions.model` 驱动每个请求，但没有任何提示词文本携带它——也不可能携带：`dsh-system-prompt` 中的 section 是上下文全局的，而模型名称因 agent（智能体）而异，`assemble()` 根本不接受任何 per-agent 输入。
|
||||
|
||||
**工具指导是 leaf YAML 中的手写行文。** bash/subagent/todo_write 的使用指导存放在 coding-agent 和 ACP（Agent Client Protocol）的 persona 字符串里——两份漂移的副本（ACP 那份已经被删减）——而 `dsh-tool-fs` 和 `dsh-tool-web` 则通过 `ctx.systemPrompt.section()` 贡献各自的指导。加载或卸载一个工具插件意味着手动编辑每个部署的 persona，旧终端欢迎横幅也手动枚举了工具集。
|
||||
**工具指导是 leaf YAML 中的手写行文。** shell/subagent/todo_write 的使用指导存放在 coding-agent 和 ACP（Agent Client Protocol）的 persona 字符串里——两份漂移的副本（ACP 那份已经被删减）——而 `dsh-tool-fs` 和 `dsh-tool-web` 则通过 `ctx.systemPrompt.section()` 贡献各自的指导。加载或卸载一个工具插件意味着手动编辑每个部署的 persona，旧终端欢迎横幅也手动枚举了工具集。
|
||||
|
||||
**Persona 渲染在工具指导之后。** agent loop（智能体循环）将 `agent.options.systemPrompt` 字符串拼接在已组装的 section 之后，于是模型先读到「Use the read tool…」再读到「You are a coding agent」——与 identity-first 约定（Claude Code、Codex）相反，且是 section 流水线之外的第二条组合路径。
|
||||
|
||||
@@ -58,7 +58,7 @@ Status: implemented
|
||||
|
||||
## 交付的不变式
|
||||
|
||||
- tui-agent 的提示词通过一条组装路径依次渲染 identity、带插值模型名的 persona，然后是 fs/bash/web 指导。
|
||||
- tui-agent 的提示词通过一条组装路径依次渲染 identity、带插值模型名的 persona，然后是 fs/shell/web 指导。
|
||||
- fork 和 fresh subagent 的描述反映提供方是否继承已完成的对话轮次；工具随提供方生命周期变化而出现、消失和重新措辞。
|
||||
- 未知、无值、格式错误或不平衡的变量引用会指明 section 名称并抛出异常；重复的 section、变量和工具注册同样抛出异常。
|
||||
- 快照回放与提示词无关：它按轮次和步骤索引已记录的分片流，不比较发出的请求。
|
||||
|
||||
@@ -2,5 +2,5 @@
|
||||
# side as of the last confirmed-consistent state. Both languages carry equal authority;
|
||||
# after editing either side, bring the other along and re-record with:
|
||||
# pnpm run verify-translation-pairing --write .agents/notes/implemented/architecture/2026-07-05-reconstructable-requests.md
|
||||
2026-07-05-reconstructable-requests.md: e78284964ca85905524d3a0800b2de46c6964094
|
||||
2026-07-05-reconstructable-requests.zh.md: 2a99fd477755c8c52f9941d832b1647786c0b242
|
||||
2026-07-05-reconstructable-requests.md: 63146fa2d392a45543daa32ce2b00158782fddb2
|
||||
2026-07-05-reconstructable-requests.zh.md: 94c1d323be0107eb8b6072a05d1e8832ebd1fffc
|
||||
|
||||
@@ -14,7 +14,7 @@ The reference shape for the happy path is MiniCode's `LLMClient`: a stateful con
|
||||
|
||||
### The principle
|
||||
|
||||
**Model-visible ⟺ durably referenced.** Anything that reaches a model request must be reconstructable from the session log and the immutable content-addressed objects it references. The checkable consequence: anyone holding the log, its referenced attachment objects, and the pinned code version reconstructs every loop request byte-for-byte. Text-only `GenerateOptions` remain a pure function of the log; image-bearing requests additionally resolve `ImageAttachmentRef` bytes through `ctx.attachments` during adapter serialization, where digest and recorded metadata verification make the object lookup deterministic and fail loud on missing or corrupt data. Direct one-shots (compaction's summarize call) log their envelope scalars (`compact/summary.{provider, model, maxTokens}`), and their input is deterministic code over the logged region plus those referenced objects — outside the invariant because only the loop marks request ownership.
|
||||
**Model-visible ⟺ durably referenced.** Anything that reaches a model request must be reconstructable from the session log and the immutable content-addressed objects it references. The checkable consequence: anyone holding the log, its referenced attachment objects, and the pinned code version reconstructs every loop request byte-for-byte. Text-only `GenerateOptions` remain a pure function of the log; image-bearing requests additionally resolve `ImageAttachmentRef` bytes through `ctx.attachments` during adapter serialization, where digest and recorded metadata verification make the object lookup deterministic and fail loud on missing or corrupt data. Direct one-shots (compaction's summarize call) log their envelope scalars (`compaction/summary.{provider, model, maxTokens}`), and their input is deterministic code over the logged region plus those referenced objects — outside the invariant because only the loop marks request ownership.
|
||||
|
||||
Prefix-cache stability is corollary #1, not the headline: an append-only log projected by a per-node pure function yields requests that are append-extensions of their predecessors whenever the header is unchanged — stability is emergent, not managed. Byte-exact audit/replay is corollary #2; resume and fork with *attributable* drift is corollary #3.
|
||||
|
||||
@@ -48,7 +48,7 @@ Like MiniCode, the conversation advances append-only and resets only when model-
|
||||
|
||||
- A request that is not explained by the log cannot be constructed by accident — not by the loop, not by a listener; mutating a built request throws; every header change is a durable, diffable log event.
|
||||
- Model-visible context uses logged message channels. `agent.inject()` and tool `additionalContexts` enter the inbox for a later claim, while `agent/pre-step` returns context that must settle with the current claimed batch. Each entered value is a durable sourced `user/message`, paid once and prefix-cached thereafter at the price of accumulating in history until compaction.
|
||||
- What still costs full price at the provider is inherent and logged: compaction (its `compact/*` events and replacement entry), a real prompt, tool, or config change (`request/header` with reason `change`), or a process boundary with drift (a differing `resume` snapshot). The provider's own reasoning-content exclusion is managed server-side.
|
||||
- What still costs full price at the provider is inherent and logged: compaction (its `compaction/*` events and replacement entry), a real prompt, tool, or config change (`request/header` with reason `change`), or a process boundary with drift (a differing `resume` snapshot). The provider's own reasoning-content exclusion is managed server-side.
|
||||
- `agent/pre-step` is the current-request message channel; direct inbox mutation is the eventual later-request channel.
|
||||
- Tool-result trimming needs no new mechanism: a logged single-entry surface replace (`start === end`) carrying a trimmed `tool/result` under the same `callId` — compaction-family, replay-correct, cache-bust batched by the same pressure logic.
|
||||
- Session logs grow one `request/header` snapshot per loop instance plus snapshots on real changes. This is larger than a delta codec but small beside chunk-heavy logs and retains one replay representation. `SESSION_FORMAT_VERSION` stays `0`; legacy delta events are rejected rather than migrated.
|
||||
|
||||
@@ -14,7 +14,7 @@ Status: implemented
|
||||
|
||||
### 原则
|
||||
|
||||
**模型可见 ⟺ 已持久引用。** 凡到达模型请求的内容都必须能从会话日志及其引用的不可变内容寻址对象中重建。可检查的推论：任何人持有日志、日志引用的附件对象和固定代码版本，即可逐字节重建循环的每个请求。纯文本 `GenerateOptions` 仍是日志的纯函数；含图片请求还会在适配器序列化期间通过 `ctx.attachments` 解析 `ImageAttachmentRef` 字节，其中对内容摘要及已记录元数据的校验使对象查找具有确定性，并在数据缺失或损坏时明确失败。直接的一次性调用（压缩的 summarize 调用）记录其信封标量（`compact/summary.{provider, model, maxTokens}`），其输入是对日志区域及这些引用对象的确定性代码运算——由于只有循环会标记请求归属，因此它们不在不变式内。
|
||||
**模型可见 ⟺ 已持久引用。** 凡到达模型请求的内容都必须能从会话日志及其引用的不可变内容寻址对象中重建。可检查的推论：任何人持有日志、日志引用的附件对象和固定代码版本，即可逐字节重建循环的每个请求。纯文本 `GenerateOptions` 仍是日志的纯函数；含图片请求还会在适配器序列化期间通过 `ctx.attachments` 解析 `ImageAttachmentRef` 字节，其中对内容摘要及已记录元数据的校验使对象查找具有确定性，并在数据缺失或损坏时明确失败。直接的一次性调用（压缩的 summarize 调用）记录其信封标量（`compaction/summary.{provider, model, maxTokens}`），其输入是对日志区域及这些引用对象的确定性代码运算——由于只有循环会标记请求归属，因此它们不在不变式内。
|
||||
|
||||
前缀缓存稳定性是推论 #1，而非标题：一个仅追加的日志经逐节点纯函数投影，在 header 不变时自然产出前一请求的追加扩展——稳定性是涌现的，不是管理出来的。字节精确的审计/回放是推论 #2；带*可归因*漂移的恢复与 fork 是推论 #3。
|
||||
|
||||
@@ -48,7 +48,7 @@ Status: implemented
|
||||
|
||||
- 一个日志无法解释的请求不可能被意外构造——无论是循环还是监听器；变异已构建的请求会抛异常；每个 header 变更都是持久的、可 diff 的日志事件。
|
||||
- 模型可见上下文使用已记录消息通道。`agent.inject()` 与工具 `additionalContexts` 进入 inbox，等待后续领取；必须与当前已领取批次一起结算的上下文由 `agent/pre-step` 返回。每个进入步骤的值都是带来源的持久 `user/message`，只付出一次代价并在后续成为可缓存前缀，代价是会在历史中累积直至压缩。
|
||||
- 在提供方处仍需全价计算的内容是固有的且已记录的：压缩（其 `compact/*` 事件和替换条目）、真正的提示词、工具或配置变更（reason 为 `change` 的 `request/header`），或带漂移的进程边界（不同的 `resume` 快照）。提供方自身的 reasoning-content 排除由服务端管理。
|
||||
- 在提供方处仍需全价计算的内容是固有的且已记录的：压缩（其 `compaction/*` 事件和替换条目）、真正的提示词、工具或配置变更（reason 为 `change` 的 `request/header`），或带漂移的进程边界（不同的 `resume` 快照）。提供方自身的 reasoning-content 排除由服务端管理。
|
||||
- `agent/pre-step` 是当前请求的消息通道；直接修改 inbox 则是最终进入后续请求的通道。
|
||||
- 工具结果裁剪无需新机制：一个已记录的单条目 surface replace（`start === end`），携带同一 `callId` 下裁剪后的 `tool/result`——属压缩家族，回放正确，缓存失效由相同的压力逻辑批量处理。
|
||||
- 会话日志每个循环实例增长一个 `request/header` 快照，并在真正变更时增加快照。它比 delta 编解码器更大，但相对分片密集型日志仍然很小，并只保留一种回放表示。`SESSION_FORMAT_VERSION` 保持 `0`；旧的 delta 事件被拒绝而非迁移。
|
||||
|
||||
@@ -2,5 +2,5 @@
|
||||
# side as of the last confirmed-consistent state. Both languages carry equal authority;
|
||||
# after editing either side, bring the other along and re-record with:
|
||||
# pnpm run verify-translation-pairing --write .agents/notes/implemented/architecture/2026-07-06-timeout-deadline-library.md
|
||||
2026-07-06-timeout-deadline-library.md: 7b252052c27fbf9f10716bd874a371b102abb614
|
||||
2026-07-06-timeout-deadline-library.zh.md: 20e5d139cceed0889adadb96abe431aef0009a2e
|
||||
2026-07-06-timeout-deadline-library.md: 38f048d16ecba0e5278ae34b0c88b7889dcfa47a
|
||||
2026-07-06-timeout-deadline-library.zh.md: e8b0ab62b19d931c33027d5e6c9919804e80aa99
|
||||
|
||||
@@ -8,8 +8,8 @@ English | [中文](2026-07-06-timeout-deadline-library.zh.md)
|
||||
|
||||
Timeout handling was drifting apart across the tool-bearing capabilities, and the divergence was not superficial — it was the same logic re-implemented three ways, each with its own subtle correctness burden.
|
||||
|
||||
- **bash** (then in the bash-local implementation's `run.ts`) had a full, correct timeout inside the process plumbing: a config-clamped `timeoutMs`, two independent triggers — a `killTimer` for the timeout and an `onAbort` listener for upstream cancellation — each calling one `kill()` closure that escalates SIGTERM→grace→SIGKILL on the process group, and two orthogonal outcome booleans (`timedOut`, `aborted`) latched independently. After this consolidation, the plumbing — today [packages/subprocess/subprocess-local/src/spawn.ts](../../../../packages/subprocess/subprocess-local/src/spawn.ts) — only reacts to aborts; [packages/bash/bash-local/src/index.ts](../../../../packages/bash/bash-local/src/index.ts) owns the fused deadline and the `timedOut`/`aborted` classification.
|
||||
- **web_fetch** ([packages/web/web-fetch-local/src/provider.ts](../../../../packages/web/web-fetch-local/src/provider.ts)) had a correct but *hand-rolled* timeout: it constructed an `AbortController`, wired `setTimeout(() => controller.abort(new WebError(…, 'WEB_FETCH_TIMEOUT')))`, manually added and removed the upstream-signal listener, cleared the timer in a `finally`, and recovered the timeout reason from `signal.reason` in a `translateAbortOrNetwork` helper because the reader surfaces a bare `AbortError`.
|
||||
- **bash** (then in the bash-local implementation's `run.ts`) had a full, correct timeout inside the process plumbing: a config-clamped `timeoutMs`, two independent triggers — a `killTimer` for the timeout and an `onAbort` listener for upstream cancellation — each calling one `kill()` closure that escalates SIGTERM→grace→SIGKILL on the process group, and two orthogonal outcome booleans (`timedOut`, `aborted`) latched independently. After this consolidation, the plumbing — today [packages/subprocess/subprocess-local/src/spawn.ts](../../../../packages/subprocess/subprocess-local/src/spawn.ts) — only reacts to aborts; [packages/shell/bash-local/src/index.ts](../../../../packages/shell/bash-local/src/index.ts) owns the fused deadline and the `timedOut`/`aborted` classification.
|
||||
- **web_fetch** ([packages/web/web-fetch-http/src/provider.ts](../../../../packages/web/web-fetch-http/src/provider.ts)) had a correct but *hand-rolled* timeout: it constructed an `AbortController`, wired `setTimeout(() => controller.abort(new WebError(…, 'WEB_FETCH_TIMEOUT')))`, manually added and removed the upstream-signal listener, cleared the timer in a `finally`, and recovered the timeout reason from `signal.reason` in a `translateAbortOrNetwork` helper because the reader surfaces a bare `AbortError`.
|
||||
- **web_search** ([packages/web/tool-web/src/search.ts](../../../../packages/web/tool-web/src/search.ts)) had **no timeout at all**: `WebSearchRequest` ([packages/web/web/src/types.ts](../../../../packages/web/web/src/types.ts)) carries no `timeoutMs` field, and each provider's `search()` only forwards `exec.signal`. (web_search stays untimed here — see Consequences.)
|
||||
|
||||
Each new external-process or network tool re-derived the same four things — clamp the requested value, start a timer, fuse the timeout with upstream cancellation, and distinguish "timed out" from "cancelled" on the way out — and the fusion and reason-recovery are exactly the parts that are easy to get subtly wrong (web_fetch's `signal.reason` dance is evidence). At the same time, the *termination* each performs is irreducibly different: bash kills an OS process group (work runs in a child process, outside this runtime, reachable only by signal), while web aborts an in-process `fetch` (undici tears down the socket). There is no single mechanism that can stop all of them.
|
||||
@@ -43,7 +43,7 @@ export function clampTimeout(
|
||||
/**
|
||||
* Build a deadline signal that aborts on upstream cancellation OR on timeout,
|
||||
* with the timeout carrying a `TimeoutReason`. `timeoutMs <= 0` means "no
|
||||
* timeout" (background tasks): forward only the upstream signal, arm no timer.
|
||||
* timeout" (background jobs): forward only the upstream signal, arm no timer.
|
||||
* The returned object's `[Symbol.dispose]` clears the timer — `using` for a
|
||||
* scope-lifetime consumer, a manual call for an event-lifetime one.
|
||||
*/
|
||||
@@ -97,13 +97,13 @@ The signal only *notifies*; termination is always the listener's job, and the li
|
||||
|
||||
## Consequences
|
||||
|
||||
- `runBash`'s outcome no longer independently latches `timedOut` and `aborted`; a timeout and a user abort racing before process close now report a single first-abort cause instead of both being true. The uniform SIGTERM→grace→SIGKILL kill is unchanged, and the Service Definition type `BashRunResult` keeps both booleans (now mutually exclusive), so `dsh-tool-bash`'s result rendering is untouched.
|
||||
- `runBash`'s outcome no longer independently latches `timedOut` and `aborted`; a timeout and a user abort racing before process close now report a single first-abort cause instead of both being true. The uniform SIGTERM→grace→SIGKILL kill is unchanged, and the Service Definition type `ShellRunResult` keeps both booleans (now mutually exclusive), so `dsh-tool-bash`'s result rendering is untouched.
|
||||
- `SpawnSpec.timeoutMs` and `SpawnOutcome.timedOut`/`aborted` were removed rather than kept as always-zero/always-false vestiges: with `runBash` owning no timer and the executor owning classification, they were read nowhere. An always-0 field read by nothing is dead weight under the per-file coverage gate.
|
||||
- web_fetch shed its bespoke controller/timer/listener/reason-recovery; the classifier now keys off the deadline signal (`timeoutOf` + `aborted`) rather than the thrown error's shape, which is robust across both the request-phase reject-with-reason and the read-phase bare-`AbortError`.
|
||||
- `AbortSignal.any` and `using`/`Symbol.dispose` enter the repo for the first time here (Node ≥ 24 baseline, already met).
|
||||
- Model streams now share one rearmable timer contract without turning a sliding idle interval into a total-call deadline or charging consumer think time. Adapters that can observe out-of-band transport activity may pulse an outstanding demand; suppressed activity remains invisible to the watchdog. The primitive still only notifies; adapter tests prove their transports observe its stable signal and terminate.
|
||||
|
||||
Out of scope, named to mark the boundary: `web_search` can gain an optional model-facing `timeout_ms` once its tool-schema/snapshot coverage is planned; the ripgrep-backed fs discovery tools ([packaged ripgrep search](2026-08-01-packaged-ripgrep-search.md)) consume the same provider-owned deadline shape through `dsh-timeout-policy` and `exec.signal`; a `tools/execute` waterfall middleware could arm a default deadline for every tool call by driving `exec.signal` — that would be a plugin that *consumes* this library and still only notifies, the hard kill remaining each capability's job.
|
||||
Out of scope, named to mark the boundary: `web_search` can gain an optional model-facing `timeout_ms` once its tool-schema/snapshot coverage is planned; the ripgrep-backed fs discovery tools ([packaged ripgrep search](2026-08-01-packaged-ripgrep-search.md)) consume the same provider-owned deadline shape through `dsh-tool-call-timeout-policy` and `exec.signal`; a `tools/execute` waterfall middleware could arm a default deadline for every tool call by driving `exec.signal` — that would be a plugin that *consumes* this library and still only notifies, the hard kill remaining each capability's job.
|
||||
|
||||
## Alternatives considered
|
||||
|
||||
|
||||
@@ -8,8 +8,8 @@ Status: implemented
|
||||
|
||||
超时处理在各个承载工具的能力之间逐渐分化，而且这种分化并非表面的：同一套逻辑被以三种方式重新实现，各自带有微妙的正确性负担。
|
||||
|
||||
- **bash**（当时位于 bash-local 实现的 `run.ts`）在进程管道内部有一套完整、正确的超时实现：一个经配置钳位的 `timeoutMs`，两个独立触发器（用于超时的 `killTimer` 和用于上游取消的 `onAbort` 监听器），各自调用同一个 `kill()` 闭包对进程组执行 SIGTERM→宽限期→SIGKILL 升级，以及两个正交的结果布尔值（`timedOut`、`aborted`）独立锁存。经此次整合之后，这套管道——今天位于 [packages/subprocess/subprocess-local/src/spawn.ts](../../../../packages/subprocess/subprocess-local/src/spawn.ts)——只响应中止；[packages/bash/bash-local/src/index.ts](../../../../packages/bash/bash-local/src/index.ts) 拥有融合的 deadline 以及 `timedOut`/`aborted` 分类。
|
||||
- **web_fetch**（[packages/web/web-fetch-local/src/provider.ts](../../../../packages/web/web-fetch-local/src/provider.ts)）有一套正确但*手写*的超时：构造一个 `AbortController`，连接 `setTimeout(() => controller.abort(new WebError(…, 'WEB_FETCH_TIMEOUT')))`，手动添加和移除上游信号监听器，在 `finally` 中清除定时器，并在 `translateAbortOrNetwork` 辅助函数中从 `signal.reason` 恢复超时原因（因为 reader 只抛出裸 `AbortError`）。
|
||||
- **bash**（当时位于 bash-local 实现的 `run.ts`）在进程管道内部有一套完整、正确的超时实现：一个经配置钳位的 `timeoutMs`，两个独立触发器（用于超时的 `killTimer` 和用于上游取消的 `onAbort` 监听器），各自调用同一个 `kill()` 闭包对进程组执行 SIGTERM→宽限期→SIGKILL 升级，以及两个正交的结果布尔值（`timedOut`、`aborted`）独立锁存。经此次整合之后，这套管道——今天位于 [packages/subprocess/subprocess-local/src/spawn.ts](../../../../packages/subprocess/subprocess-local/src/spawn.ts)——只响应中止；[packages/shell/bash-local/src/index.ts](../../../../packages/shell/bash-local/src/index.ts) 拥有融合的 deadline 以及 `timedOut`/`aborted` 分类。
|
||||
- **web_fetch**（[packages/web/web-fetch-http/src/provider.ts](../../../../packages/web/web-fetch-http/src/provider.ts)）有一套正确但*手写*的超时：构造一个 `AbortController`，连接 `setTimeout(() => controller.abort(new WebError(…, 'WEB_FETCH_TIMEOUT')))`，手动添加和移除上游信号监听器，在 `finally` 中清除定时器，并在 `translateAbortOrNetwork` 辅助函数中从 `signal.reason` 恢复超时原因（因为 reader 只抛出裸 `AbortError`）。
|
||||
- **web_search**（[packages/web/tool-web/src/search.ts](../../../../packages/web/tool-web/src/search.ts)）**完全没有超时**：`WebSearchRequest`（[packages/web/web/src/types.ts](../../../../packages/web/web/src/types.ts)）不携带 `timeoutMs` 字段，各提供方的 `search()` 只转发 `exec.signal`。（web_search 在本次设计中保持无超时——见「后果」。）
|
||||
|
||||
每个新的外部进程或网络工具都要重新推导同样四件事：钳位请求值、启动定时器、将超时与上游取消融合、在出口处区分「超时」与「已取消」。而融合与原因恢复恰恰是最容易出微妙错误的部分（web_fetch 的 `signal.reason` 处理就是证据）。与此同时，各能力执行的*终止*操作不可归约地不同：bash 杀死一个 OS 进程组（工作运行在子进程中，在本运行时之外，只能通过信号触达），而 web 中止一个进程内的 `fetch`（undici 拆除 socket）。不存在一个能停止所有能力工作的单一机制。
|
||||
@@ -43,7 +43,7 @@ export function clampTimeout(
|
||||
/**
|
||||
* Build a deadline signal that aborts on upstream cancellation OR on timeout,
|
||||
* with the timeout carrying a `TimeoutReason`. `timeoutMs <= 0` means "no
|
||||
* timeout" (background tasks): forward only the upstream signal, arm no timer.
|
||||
* timeout" (background jobs): forward only the upstream signal, arm no timer.
|
||||
* The returned object's `[Symbol.dispose]` clears the timer — `using` for a
|
||||
* scope-lifetime consumer, a manual call for an event-lifetime one.
|
||||
*/
|
||||
@@ -97,13 +97,13 @@ export function timeoutOf(x: AbortSignal | { reason?: unknown }, code?: string):
|
||||
|
||||
## 后果
|
||||
|
||||
- `runBash` 的结果不再独立锁存 `timedOut` 和 `aborted`；超时与用户中止在进程关闭前竞争时，现在报告单一的首个 abort 原因，而非两者同时为 true。统一的 SIGTERM→宽限期→SIGKILL 终止路径不变，Service Definition 类型 `BashRunResult` 保留两个布尔值（现在互斥），因此 `dsh-tool-bash` 的结果渲染不受影响。
|
||||
- `runBash` 的结果不再独立锁存 `timedOut` 和 `aborted`；超时与用户中止在进程关闭前竞争时，现在报告单一的首个 abort 原因，而非两者同时为 true。统一的 SIGTERM→宽限期→SIGKILL 终止路径不变，Service Definition 类型 `ShellRunResult` 保留两个布尔值（现在互斥），因此 `dsh-tool-bash` 的结果渲染不受影响。
|
||||
- `SpawnSpec.timeoutMs` 和 `SpawnOutcome.timedOut`/`aborted` 被移除，而非作为始终为零/始终为 false 的残余保留：由于 `runBash` 不再拥有定时器且执行器负责分类，这些字段无处被读取。一个始终为 0 且无处读取的字段在逐文件覆盖率门禁下属于死代码。
|
||||
- web_fetch 去除了其定制的 controller/timer/listener/reason-recovery；分类器现在基于 deadline 信号（`timeoutOf` + `aborted`）而非抛出错误的形状来判断，这在请求阶段的 reject-with-reason 和读取阶段的裸 `AbortError` 两种情况下都是健壮的。
|
||||
- `AbortSignal.any` 和 `using`/`Symbol.dispose` 在此首次进入本仓库（Node ≥ 24 基线，已满足）。
|
||||
- 模型流现在共享一个可重启的定时器约定，不会把滑动的空闲间隔变成总调用截止时间，也不会计入消费方思考时间。能够观察到带外传输活动的适配器可以对尚未结算的 demand 调用 `pulse()`；被屏蔽的活动对 watchdog 仍不可见。该原语仍然只做通知；适配器测试证明其传输观察到稳定信号并终止。
|
||||
|
||||
以下内容不在本次范围内，列出以标明边界：`web_search` 可以在其工具 schema 和快照覆盖规划完成后获得可选的面向模型的 `timeout_ms`；基于 ripgrep 的文件系统发现工具（[打包的 ripgrep 搜索](2026-08-01-packaged-ripgrep-search.md)）通过 `dsh-timeout-policy` 和 `exec.signal` 消费同样的提供方自有 deadline 形状；`tools/execute` waterfall（瀑布式事件）中间件可以通过驱动 `exec.signal` 为每次工具调用设置默认 deadline——那将是一个*消费*本库的插件，仍然只做通知，硬终止仍是各能力自己的事。
|
||||
以下内容不在本次范围内，列出以标明边界：`web_search` 可以在其工具 schema 和快照覆盖规划完成后获得可选的面向模型的 `timeout_ms`；基于 ripgrep 的文件系统发现工具（[打包的 ripgrep 搜索](2026-08-01-packaged-ripgrep-search.md)）通过 `dsh-tool-call-timeout-policy` 和 `exec.signal` 消费同样的提供方自有 deadline 形状；`tools/execute` waterfall（瀑布式事件）中间件可以通过驱动 `exec.signal` 为每次工具调用设置默认 deadline——那将是一个*消费*本库的插件，仍然只做通知，硬终止仍是各能力自己的事。
|
||||
|
||||
## 曾考虑的替代方案
|
||||
|
||||
|
||||
@@ -2,5 +2,5 @@
|
||||
# side as of the last confirmed-consistent state. Both languages carry equal authority;
|
||||
# after editing either side, bring the other along and re-record with:
|
||||
# pnpm run verify-translation-pairing --write .agents/notes/implemented/architecture/2026-07-06-tool-result-retention-library.md
|
||||
2026-07-06-tool-result-retention-library.md: 1736d2dad98cbb8b67d570ce25742a84ea0ede59
|
||||
2026-07-06-tool-result-retention-library.zh.md: f5efc5dc5efccea8ca0f5b8b4cadbd251e5a7fe7
|
||||
2026-07-06-tool-result-retention-library.md: 8d938db8f5fa78398a39e97cc877308200f7d60f
|
||||
2026-07-06-tool-result-retention-library.zh.md: 0b2fe841beed21511c4088732713d3fd9fd9048f
|
||||
|
||||
@@ -12,7 +12,7 @@ The shared abstraction the tools need is **retention**, not generic collection.
|
||||
|
||||
## Decision
|
||||
|
||||
`@deepseek-ai/dsh-retention` lives under `packages/util/` (peer to `dsh-brand` and `dsh-timeout`) and owns bounded model-facing output. It is a library of pure classes and functions, **not** a Cordis service or plugin: it takes no `ctx`, registers nothing, holds no cross-call state, and emits no events. Tool packages import it directly when they need bounded output.
|
||||
`@deepseek-ai/dsh-output-retention` lives under `packages/util/` (peer to `dsh-brand` and `dsh-timeout`) and owns bounded model-facing output. It is a library of pure classes and functions, **not** a Cordis service or plugin: it takes no `ctx`, registers nothing, holds no cross-call state, and emits no events. Tool packages import it directly when they need bounded output.
|
||||
|
||||
The library has two independent retainers:
|
||||
|
||||
@@ -103,7 +103,7 @@ type TextRetentionStrategy =
|
||||
|
||||
`grep` uses `ItemRetainer<FlatGrepMatch>` with `{ kind: 'head', maxItems: grepMaxMatches }` before grouping. The executor parses ripgrep output, maps paths, applies per-line preview truncation, and pushes flat matches. After `finish()`, the tool groups retained matches by file and can save the full match list through the spill seam when the inline result is capped. Grouping is not part of the retainer because the cap is total matches, not files; per-match preview truncation and `incomplete` are also separate from result-level retention.
|
||||
|
||||
`bash` can use `TextRetainer` with `tail` or `headTail` and reads to process completion. The bash executor still owns spill files, exit status, signal, timeout, and background-task behavior; the retention helper only replaces ad hoc in-memory head/tail accounting where that behavior is desired. Long-running task ownership remains orthogonal to the [generic long-running tool runtime](2026-06-20-generic-long-running-tool-runtime.md).
|
||||
`bash` can use `TextRetainer` with `tail` or `headTail` and reads to process completion. The bash executor still owns spill files, exit status, signal, timeout, and background-job behavior; the retention helper only replaces ad hoc in-memory head/tail accounting where that behavior is desired. Long-running job ownership remains orthogonal to the [generic long-running tool runtime](2026-06-20-generic-long-running-tool-runtime.md).
|
||||
|
||||
`web_fetch` can use `TextRetainer` with `head` or `headTail`, or keep provider-owned body caps when the provider must read and decode internally. Either way, the fetch result's `truncated` remains a provider/tool fact, and the library only supplies retained text and omission metadata.
|
||||
|
||||
@@ -136,9 +136,9 @@ The formatter hook is deliberately small: a tool turns a `RetentionNotice` into
|
||||
|
||||
## Consequences
|
||||
|
||||
**What shipped.** `@deepseek-ai/dsh-retention` exports `ItemRetainer`, `TextRetainer`, the result types (`RetainedItems`, `RetainedText`), the strategy types (`ItemRetentionStrategy`, `TextRetentionStrategy`), `Omitted`, `PushDecision`, `RetentionNotice`, and the neutral notice helpers `describeOmitted` / `formatRetentionNotice` — with no dependency on Cordis or any tool package. Unit tests cover item-head retention with exact omission counts, text-head retention, text-tail retention, head-tail byte retention, zero budgets, UTF-8 boundary handling (2-, 3-, and 4-byte codepoints and invalid lead bytes at each cut), and unknown omission wording.
|
||||
**What shipped.** `@deepseek-ai/dsh-output-retention` exports `ItemRetainer`, `TextRetainer`, the result types (`RetainedItems`, `RetainedText`), the strategy types (`ItemRetentionStrategy`, `TextRetentionStrategy`), `Omitted`, `PushDecision`, `RetentionNotice`, and the neutral notice helpers `describeOmitted` / `formatRetentionNotice` — with no dependency on Cordis or any tool package. Unit tests cover item-head retention with exact omission counts, text-head retention, text-tail retention, head-tail byte retention, zero budgets, UTF-8 boundary handling (2-, 3-, and 4-byte codepoints and invalid lead bytes at each cut), and unknown omission wording.
|
||||
|
||||
**What is documented but not yet migrated.** `glob`, `grep`, `bash`, `web_fetch`, and `web_search` have their mappings documented in the [package README](../../../../packages/util/retention/README.md), but not every tool has been migrated onto the library in this change; migration is deliberately separate follow-up work. `read` is documented as intentionally out of scope: its `read-render` line-window contract (`offset`/`limit`, `totalLines`, offset-range errors, per-line preview truncation, a byte cap over the selected window) is not generic retention, and one `Omitted` count cannot represent both sides of a line window.
|
||||
**What is documented but not yet migrated.** `glob`, `grep`, `bash`, `web_fetch`, and `web_search` have their mappings documented in the [package README](../../../../packages/util/output-retention/README.md), but not every tool has been migrated onto the library in this change; migration is deliberately separate follow-up work. `read` is documented as intentionally out of scope: its `read-render` line-window contract (`offset`/`limit`, `totalLines`, offset-range errors, per-line preview truncation, a byte cap over the selected window) is not generic retention, and one `Omitted` count cannot represent both sides of a line window.
|
||||
|
||||
**Boundaries the library holds.** `truncated` means the retainer omitted otherwise-available content because of a budget; it never means the upstream was incomplete. Tool-specific states — `incomplete`, permission failures, provider partial failures, binary skips, bash spill-path recovery, invalid UTF-8 — stay in tool-domain fields, outside the retainer. When a future change migrates a tool, that package's README and tests must prove the model-facing result text is unchanged except for deliberate notice wording.
|
||||
|
||||
|
||||
@@ -12,7 +12,7 @@ Status: implemented
|
||||
|
||||
## 决策
|
||||
|
||||
`@deepseek-ai/dsh-retention` 位于 `packages/util/` 下，与 `dsh-brand` 和 `dsh-timeout` 同级，负责有界的模型可见输出。它是一组纯类与函数构成的库，**不是** Cordis 服务或插件：不接收 `ctx`、不注册任何内容、不持有跨调用状态，也不发出事件。各工具包需要限制输出时直接导入它。
|
||||
`@deepseek-ai/dsh-output-retention` 位于 `packages/util/` 下，与 `dsh-brand` 和 `dsh-timeout` 同级，负责有界的模型可见输出。它是一组纯类与函数构成的库，**不是** Cordis 服务或插件：不接收 `ctx`、不注册任何内容、不持有跨调用状态，也不发出事件。各工具包需要限制输出时直接导入它。
|
||||
|
||||
该库包含两个相互独立的 retainer：
|
||||
|
||||
@@ -136,9 +136,9 @@ const formatGrepNotice = (notice: RetentionNotice): string =>
|
||||
|
||||
## 影响
|
||||
|
||||
**已交付内容。** `@deepseek-ai/dsh-retention` 导出 `ItemRetainer`、`TextRetainer`、结果类型（`RetainedItems`、`RetainedText`）、策略类型（`ItemRetentionStrategy`、`TextRetentionStrategy`）、`Omitted`、`PushDecision`、`RetentionNotice`，以及中性的提示辅助函数 `describeOmitted`／`formatRetentionNotice`，且不依赖 Cordis 或任何工具包。单元测试覆盖具有精确省略计数的条目头部保留、文本头部保留、文本尾部保留、首尾字节保留、零预算、UTF-8 边界处理（2、3、4 字节码位，以及每个裁切位置上的无效起始字节）和未知省略量的措辞。
|
||||
**已交付内容。** `@deepseek-ai/dsh-output-retention` 导出 `ItemRetainer`、`TextRetainer`、结果类型（`RetainedItems`、`RetainedText`）、策略类型（`ItemRetentionStrategy`、`TextRetentionStrategy`）、`Omitted`、`PushDecision`、`RetentionNotice`，以及中性的提示辅助函数 `describeOmitted`／`formatRetentionNotice`，且不依赖 Cordis 或任何工具包。单元测试覆盖具有精确省略计数的条目头部保留、文本头部保留、文本尾部保留、首尾字节保留、零预算、UTF-8 边界处理（2、3、4 字节码位，以及每个裁切位置上的无效起始字节）和未知省略量的措辞。
|
||||
|
||||
**已记录但尚未迁移的内容。** `glob`、`grep`、`bash`、`web_fetch` 与 `web_search` 的映射已记录在[包 README](../../../../packages/util/retention/README.md) 中，但本次改动并未把每个工具都迁移到该库；迁移工作刻意留作独立的后续任务。`read` 被明确记录为不在范围内：其 `read-render` 行窗口约定（`offset`／`limit`、`totalLines`、offset 范围错误、逐行预览截断，以及针对所选窗口的字节上限）不属于通用保留，而一个 `Omitted` 计数也无法同时表达行窗口两侧。
|
||||
**已记录但尚未迁移的内容。** `glob`、`grep`、`bash`、`web_fetch` 与 `web_search` 的映射已记录在[包 README](../../../../packages/util/output-retention/README.md) 中，但本次改动并未把每个工具都迁移到该库；迁移工作刻意留作独立的后续任务。`read` 被明确记录为不在范围内：其 `read-render` 行窗口约定（`offset`／`limit`、`totalLines`、offset 范围错误、逐行预览截断，以及针对所选窗口的字节上限）不属于通用保留，而一个 `Omitted` 计数也无法同时表达行窗口两侧。
|
||||
|
||||
**该库维持的边界。** `truncated` 表示 retainer 因预算省略了原本可用的内容，绝不表示上游不完整。工具专用状态，包括 `incomplete`、权限失败、提供方局部失败、跳过二进制文件、bash spill 路径恢复和无效 UTF-8，均留在工具领域字段中、位于 retainer 之外。未来改动迁移某项工具时，该包的 README 与测试必须证明，除了有意改变的提示措辞外，模型可见的结果文本没有变化。
|
||||
|
||||
|
||||
@@ -2,5 +2,5 @@
|
||||
# side as of the last confirmed-consistent state. Both languages carry equal authority;
|
||||
# after editing either side, bring the other along and re-record with:
|
||||
# pnpm run verify-translation-pairing --write .agents/notes/implemented/architecture/2026-07-07-tool-call-timeout-policy.md
|
||||
2026-07-07-tool-call-timeout-policy.md: f17834332b2e661b6811c2264d5e8de4e1642663
|
||||
2026-07-07-tool-call-timeout-policy.zh.md: b5da0aa63ecd83756d0ce136317088d42bfd65cf
|
||||
2026-07-07-tool-call-timeout-policy.md: 20ca57cee4cd3ae51e7bfb8af1d2a1f8c7edd743
|
||||
2026-07-07-tool-call-timeout-policy.zh.md: 5b91c29f5492a36bbd023a892f0f2f8dd857ae53
|
||||
|
||||
@@ -8,7 +8,7 @@ English | [中文](2026-07-07-tool-call-timeout-policy.zh.md)
|
||||
|
||||
The [timeout/deadline Agent Note](2026-07-06-timeout-deadline-library.md) extracted the timing-and-classification primitive into `@deepseek-ai/dsh-timeout`, but timeout policy was still attached to individual capabilities and model-facing schemas. `bash` exposed `timeoutMs`; `web_fetch` exposed `timeout_ms`; `web_search` had no model-facing timeout even though providers already honor `exec.signal`; a future grep/glob tool would either import the timeout library directly or invent its own timeout policy. That is the wrong authoring shape for a plugin SDK: a tool author should normally forward `exec.signal` to the implementation it calls, and deployment policy should decide the budget.
|
||||
|
||||
At the same time, not every timeout in the repo is a model-facing tool-call budget. Hooks execute command hooks by calling `ctx.bash` directly, not through `ctx.tools.execute()`, and the `bash` model tool multiplexes foreground execution, background start, background polling, and hook reuse through the same backend. Moving every timeout into a tool plugin in one step would conflate those paths and risk breaking hook timeout semantics.
|
||||
At the same time, not every timeout in the repo is a model-facing tool-call budget. Hooks execute command hooks by calling `ctx.shell` directly, not through `ctx.tools.execute()`, and the `bash` model tool multiplexes foreground execution, background start, background polling, and hook reuse through the same backend. Moving every timeout into a tool plugin in one step would conflate those paths and risk breaking hook timeout semantics.
|
||||
|
||||
## Decision
|
||||
|
||||
@@ -16,7 +16,7 @@ Tool-call timeout is a policy that applies only to model-facing tool execution,
|
||||
|
||||
- `@deepseek-ai/dsh-timeout` remains the shared library that owns `deadline()` and `timeoutOf()`.
|
||||
- `@deepseek-ai/dsh-tools` has an around-dispatch waterfall, `tools/execute`, between `tools/pre-execute` and `tools/post-execute`.
|
||||
- `@deepseek-ai/dsh-timeout-policy` reads each tool's declared `timeoutMs` from the registry and wraps a call that has one by deriving a new `exec.signal`.
|
||||
- The [repository naming contract](../../proposed/architecture/2026-08-11-repository-naming-contract-and-rename-ledger.md) names `@deepseek-ai/dsh-tool-call-timeout-policy` for the exact operation it limits. The plugin reads each tool's declared `timeoutMs` from the runtime and wraps a call that has one by deriving a new `exec.signal`.
|
||||
|
||||
The execution pipeline is:
|
||||
|
||||
@@ -40,11 +40,11 @@ That the catch is the base `next` — not something outside the waterfall — is
|
||||
|
||||
### The `timeout-policy` plugin
|
||||
|
||||
The plugin is `@deepseek-ai/dsh-timeout-policy`, a zero-config function/namespace plugin (`name` / `inject` / `apply`) in the `packages/guard/` group (originally its own `timeout/` group). The per-tool budget is DECLARED on the tool, not on this plugin: a `ToolDefinition` carries an optional `timeoutMs`, which the owning tool plugin sets from its own config. `dsh-tool-web`, for example, resolves `fetchTimeoutMs` / `searchTimeoutMs` (default 30000) onto the `web_fetch` / `web_search` definitions:
|
||||
The plugin is `@deepseek-ai/dsh-tool-call-timeout-policy`, a zero-config function/namespace plugin (`name` / `inject` / `apply`) in the `packages/guard/` group (originally its own `timeout/` group). The per-tool budget is DECLARED on the tool, not on this plugin: a `ToolDefinition` carries an optional `timeoutMs`, which the owning tool plugin sets from its own config. `dsh-tool-web`, for example, resolves `fetchTimeoutMs` / `searchTimeoutMs` (default 30000) onto the `web_fetch` / `web_search` definitions:
|
||||
|
||||
```yaml
|
||||
- id: timeout-policy
|
||||
name: '@deepseek-ai/dsh-timeout-policy'
|
||||
name: '@deepseek-ai/dsh-tool-call-timeout-policy'
|
||||
- id: tool-web
|
||||
name: '@deepseek-ai/dsh-tool-web'
|
||||
config:
|
||||
@@ -79,21 +79,21 @@ No new session event is needed for reconstructability: `TOOL_TIMEOUT` is the fin
|
||||
|
||||
`web_fetch` and `web_search` are migrated. `dsh-tool-web` keeps ownership of their model-facing schemas, and those schemas expose no timeout knob: `web_fetch` dropped its `timeout_ms` parameter to match the reference-agent shape, and `web_search` stays query-only. The tool bodies do not import `@deepseek-ai/dsh-timeout`; they forward `exec.signal` to `ctx.web`.
|
||||
|
||||
`dsh-web-fetch-local` keeps one configured provider-level `timeoutMs` as a large resource backstop for direct `ctx.web.fetch()` callers and misconfigured deployments; it owns no model-facing timeout. When a `TOOL_TIMEOUT` signal reaches the fetch provider first, provider-scoped classification treats it as upstream `WEB_ABORTED`, and the outer `tools/execute` wrapper replaces the final tool result with `TOOL_TIMEOUT`. A shipped web-tool deployment configures the provider backstop above the `timeout-policy` budget so the tool-call policy normally wins for model calls.
|
||||
`dsh-web-fetch-http` keeps one configured provider-level `timeoutMs` as a large resource backstop for direct `ctx.web.fetch()` callers and misconfigured deployments; it owns no model-facing timeout. When a `TOOL_TIMEOUT` signal reaches the fetch provider first, provider-scoped classification treats it as upstream `WEB_ABORTED`, and the outer `tools/execute` wrapper replaces the final tool result with `TOOL_TIMEOUT`. A shipped web-tool deployment configures the provider backstop above the `timeout-policy` budget so the tool-call policy normally wins for model calls.
|
||||
|
||||
`bash` stays on the current backend timeout path. `dsh-tool-bash` continues to expose `timeoutMs` and `run_in_background`; `dsh-bash-local` continues to use `@deepseek-ai/dsh-timeout` for `BASH_TIMEOUT`; hook bridges continue to call `runHook()` and pass `timeoutMs` through `ctx.bash`. This keeps foreground/background/hook behavior stable.
|
||||
`bash` stays on the current backend timeout path. `dsh-tool-bash` continues to expose `timeoutMs` and `run_in_background`; `dsh-bash-local` continues to use `@deepseek-ai/dsh-timeout` for `BASH_TIMEOUT`; hook bridges continue to call `runHook()` and pass `timeoutMs` through `ctx.shell`. This keeps foreground/background/hook behavior stable.
|
||||
|
||||
`read`, `write`, `edit`, `todo_write`, `task_list`, and `task_kill` do not opt into tool-call timeout. `task_output` owns its bounded wait because a wait timeout is a successful live-status result, not a tool failure.
|
||||
`read`, `write`, `edit`, `todo_write`, `job_list`, and `job_kill` do not opt into tool-call timeout. `job_output` owns its bounded wait because a wait timeout is a successful live-status result, not a tool failure.
|
||||
|
||||
A future model-facing grep/glob tool can be implemented on top of `ctx.bash` without importing `@deepseek-ai/dsh-timeout`: it forwards `exec.signal` to `ctx.bash`, and declares its own `timeoutMs` (from its plugin's config) for the enforcer to apply. If bash-local's backend timeout becomes a problem for such a tool, the bash seam can later add a caller-owned-deadline mode; that is a separate decision.
|
||||
A future model-facing grep/glob tool can be implemented on top of `ctx.shell` without importing `@deepseek-ai/dsh-timeout`: it forwards `exec.signal` to `ctx.shell`, and declares its own `timeoutMs` (from its plugin's config) for the enforcer to apply. If bash-local's backend timeout becomes a problem for such a tool, the bash seam can later add a caller-owned-deadline mode; that is a separate decision.
|
||||
|
||||
## Alternatives considered
|
||||
|
||||
**Name the plugin `tool-timeout`.** The literal Agent Note name matched the `gen-tool-catalog` completeness guard's `packages/*/tool-*` glob, which requires every match to register a model-facing tool. This plugin registers none — it is a `tools/execute` wrapper — so a `tool-*` name would either fail `verify-tool-catalog` or force a misleading boot entry. The package is `@deepseek-ai/dsh-timeout-policy` in what was then a new `timeout/` group, since folded into `packages/guard/`; the cordis.yml `id` can still be `timeout-policy`.
|
||||
**Name the plugin `tool-timeout`.** The literal Agent Note name matched the `gen-tool-catalog` completeness guard's `packages/*/tool-*` glob, which requires every match to register a model-facing tool. This plugin registers none — it is a `tools/execute` wrapper — so a `tool-*` name would either fail `verify-tool-catalog` or force a misleading boot entry. The package is `@deepseek-ai/dsh-tool-call-timeout-policy` in what was then a new `timeout/` group, since folded into `packages/guard/`; the cordis.yml `id` can still be `timeout-policy`.
|
||||
|
||||
**Keep per-tool timeout handling only.** This was the shape for `bash` and `web_fetch`, and it matches Claude Code and Codex for shell commands. It loses for web-style tools because every new timeout-capable tool must choose validation, cap semantics, docs, snapshots, and classification. The plugin centralizes policy and classification while leaving each tool's schema focused on business input.
|
||||
|
||||
**Move all timeout policy out of bash-local immediately.** Cleaner long-term — bash-local would become a pure subprocess executor and all callers would own their deadlines. It loses as the first step because hooks call `ctx.bash` directly and the bash model tool has foreground/background semantics that are not the same tool-call lifetime. Keeping `BASH_TIMEOUT` preserves those paths while tool-call timeout proves itself on simpler tools.
|
||||
**Move all timeout policy out of bash-local immediately.** Cleaner long-term — bash-local would become a pure subprocess executor and all callers would own their deadlines. It loses as the first step because hooks call `ctx.shell` directly and the bash model tool has foreground/background semantics that are not the same tool-call lifetime. Keeping `BASH_TIMEOUT` preserves those paths while tool-call timeout proves itself on simpler tools.
|
||||
|
||||
**Use a global default budget for every tool.** Convenient, but it surprises tool authors: any tool that accidentally runs longer than the global budget would start failing once the plugin loads. A per-tool declared budget makes adoption deliberate.
|
||||
|
||||
|
||||
@@ -8,7 +8,7 @@ Status: implemented
|
||||
|
||||
[超时/截止时间 Agent Note](2026-07-06-timeout-deadline-library.md) 将计时与分类原语提取到了 `@deepseek-ai/dsh-timeout`，但超时策略仍然附着在各个能力和面向模型的 schema 上。`bash` 暴露了 `timeoutMs`；`web_fetch` 暴露了 `timeout_ms`；`web_search` 没有面向模型的超时参数，尽管提供方已经遵循 `exec.signal`；未来的 grep/glob 工具要么直接导入超时库，要么自行发明超时策略。对于一个插件 SDK 来说，这是错误的编写范式：工具作者通常只需将 `exec.signal` 转发给其调用的实现，而部署策略来决定预算。
|
||||
|
||||
与此同时，仓库中并非所有超时都是面向模型的工具调用预算。钩子通过直接调用 `ctx.bash` 执行命令钩子，而非通过 `ctx.tools.execute()`；`bash` 模型工具通过同一个后端复用前台执行、后台启动、后台轮询和钩子复用。一步到位地将所有超时移入工具插件会混淆这些路径，并有破坏钩子超时语义的风险。
|
||||
与此同时，仓库中并非所有超时都是面向模型的工具调用预算。钩子通过直接调用 `ctx.shell` 执行命令钩子，而非通过 `ctx.tools.execute()`；`bash` 模型工具通过同一个后端复用前台执行、后台启动、后台轮询和钩子复用。一步到位地将所有超时移入工具插件会混淆这些路径，并有破坏钩子超时语义的风险。
|
||||
|
||||
## 决策
|
||||
|
||||
@@ -16,7 +16,7 @@ Status: implemented
|
||||
|
||||
- `@deepseek-ai/dsh-timeout` 仍是拥有 `deadline()` 和 `timeoutOf()` 的共享库。
|
||||
- `@deepseek-ai/dsh-tools` 在 `tools/pre-execute` 和 `tools/post-execute` 之间有一个环绕分发的 waterfall（瀑布式事件）`tools/execute`。
|
||||
- `@deepseek-ai/dsh-timeout-policy` 从注册表读取每个工具声明的 `timeoutMs`，并通过派生新的 `exec.signal` 来包装有此声明的调用。
|
||||
- [仓库命名约定](../../proposed/architecture/2026-08-11-repository-naming-contract-and-rename-ledger.md)使用 `@deepseek-ai/dsh-tool-call-timeout-policy`，准确说明该策略所限制的操作。插件从 runtime 读取每个工具声明的 `timeoutMs`，并通过派生新的 `exec.signal` 来包装有此声明的调用。
|
||||
|
||||
执行流水线如下：
|
||||
|
||||
@@ -40,11 +40,11 @@ catch 是基础 `next`（而非 waterfall 之外的东西）这一点至关重
|
||||
|
||||
### `timeout-policy` 插件
|
||||
|
||||
该插件是 `@deepseek-ai/dsh-timeout-policy`，一个零配置的函数/命名空间插件（`name` / `inject` / `apply`），位于 `packages/guard/` 组。每个工具的预算声明在工具自身，而非本插件：`ToolDefinition` 携带一个可选的 `timeoutMs`，由拥有该工具的插件从自身配置中设置。例如 `dsh-tool-web` 将 `fetchTimeoutMs` / `searchTimeoutMs`（默认 30000）解析到 `web_fetch` / `web_search` 的定义上：
|
||||
该插件是 `@deepseek-ai/dsh-tool-call-timeout-policy`，一个零配置的函数/命名空间插件（`name` / `inject` / `apply`），位于 `packages/guard/` 组。每个工具的预算声明在工具自身，而非本插件：`ToolDefinition` 携带一个可选的 `timeoutMs`，由拥有该工具的插件从自身配置中设置。例如 `dsh-tool-web` 将 `fetchTimeoutMs` / `searchTimeoutMs`（默认 30000）解析到 `web_fetch` / `web_search` 的定义上：
|
||||
|
||||
```yaml
|
||||
- id: timeout-policy
|
||||
name: '@deepseek-ai/dsh-timeout-policy'
|
||||
name: '@deepseek-ai/dsh-tool-call-timeout-policy'
|
||||
- id: tool-web
|
||||
name: '@deepseek-ai/dsh-tool-web'
|
||||
config:
|
||||
@@ -79,21 +79,21 @@ function toolTimeoutResult(timeoutMs: number): ToolExecutionResult {
|
||||
|
||||
`web_fetch` 和 `web_search` 已迁移。`dsh-tool-web` 保留对其面向模型 schema 的所有权，这些 schema 不暴露超时旋钮：`web_fetch` 移除了 `timeout_ms` 参数以匹配参考 agent（智能体）的形状，`web_search` 保持仅查询。工具体不导入 `@deepseek-ai/dsh-timeout`；它们将 `exec.signal` 转发给 `ctx.web`。
|
||||
|
||||
`dsh-web-fetch-local` 保留一个在提供方层面配置的 `timeoutMs`，作为较大的资源兜底值，服务于直接调用 `ctx.web.fetch()` 的调用方和配置错误的部署；它不拥有面向模型的超时。当 `TOOL_TIMEOUT` 信号先到达 fetch 提供方时，提供方作用域的分类将其视为上游 `WEB_ABORTED`，而外层 `tools/execute` 包装器将最终工具结果替换为 `TOOL_TIMEOUT`。一个已发布的 web 工具部署将提供方兜底配置为高于 `timeout-policy` 预算，使工具调用策略在模型调用中通常胜出。
|
||||
`dsh-web-fetch-http` 保留一个在提供方层面配置的 `timeoutMs`，作为较大的资源兜底值，服务于直接调用 `ctx.web.fetch()` 的调用方和配置错误的部署；它不拥有面向模型的超时。当 `TOOL_TIMEOUT` 信号先到达 fetch 提供方时，提供方作用域的分类将其视为上游 `WEB_ABORTED`，而外层 `tools/execute` 包装器将最终工具结果替换为 `TOOL_TIMEOUT`。一个已发布的 web 工具部署将提供方兜底配置为高于 `timeout-policy` 预算，使工具调用策略在模型调用中通常胜出。
|
||||
|
||||
`bash` 保持当前的后端超时路径。`dsh-tool-bash` 继续暴露 `timeoutMs` 和 `run_in_background`；`dsh-bash-local` 继续使用 `@deepseek-ai/dsh-timeout` 处理 `BASH_TIMEOUT`；钩子桥接继续调用 `runHook()` 并通过 `ctx.bash` 传递 `timeoutMs`。这保持了前台/后台/钩子行为的稳定。
|
||||
`bash` 保持当前的后端超时路径。`dsh-tool-bash` 继续暴露 `timeoutMs` 和 `run_in_background`；`dsh-bash-local` 继续使用 `@deepseek-ai/dsh-timeout` 处理 `BASH_TIMEOUT`；钩子桥接继续调用 `runHook()` 并通过 `ctx.shell` 传递 `timeoutMs`。这保持了前台/后台/钩子行为的稳定。
|
||||
|
||||
`read`、`write`、`edit`、`todo_write`、`task_list` 和 `task_kill` 不加入工具调用超时。`task_output` 自己拥有有界等待，因为等待超时是成功的实时状态结果，而非工具失败。
|
||||
`read`、`write`、`edit`、`todo_write`、`job_list` 和 `job_kill` 不加入工具调用超时。`job_output` 自己拥有有界等待，因为等待超时是成功的实时状态结果，而非工具失败。
|
||||
|
||||
未来面向模型的 grep/glob 工具可以基于 `ctx.bash` 实现而无需导入 `@deepseek-ai/dsh-timeout`：它将 `exec.signal` 转发给 `ctx.bash`，并声明自己的 `timeoutMs`（来自其插件配置）供执行器应用。如果 bash-local 的后端超时对这类工具造成问题，bash seam 可以后续添加调用方自有截止模式；那是一项独立的决策。
|
||||
未来面向模型的 grep/glob 工具可以基于 `ctx.shell` 实现而无需导入 `@deepseek-ai/dsh-timeout`：它将 `exec.signal` 转发给 `ctx.shell`，并声明自己的 `timeoutMs`（来自其插件配置）供执行器应用。如果 bash-local 的后端超时对这类工具造成问题，bash seam 可以后续添加调用方自有截止模式；那是一项独立的决策。
|
||||
|
||||
## 曾考虑的替代方案
|
||||
|
||||
**将插件命名为 `tool-timeout`。** 字面的 Agent Note 名称匹配了 `gen-tool-catalog` 完整性守卫的 `packages/*/tool-*` glob，该 glob 要求每个匹配项注册一个面向模型的工具。本插件不注册任何工具——它是一个 `tools/execute` 包装器——因此 `tool-*` 名称要么导致 `verify-tool-catalog` 失败，要么强制产生一个误导性的启动条目。包为 `@deepseek-ai/dsh-timeout-policy`，位于新的 `packages/guard/` 组；cordis.yml 的 `id` 仍可为 `timeout-policy`。
|
||||
**将插件命名为 `tool-timeout`。** 字面的 Agent Note 名称匹配了 `gen-tool-catalog` 完整性守卫的 `packages/*/tool-*` glob，该 glob 要求每个匹配项注册一个面向模型的工具。本插件不注册任何工具——它是一个 `tools/execute` 包装器——因此 `tool-*` 名称要么导致 `verify-tool-catalog` 失败，要么强制产生一个误导性的启动条目。包为 `@deepseek-ai/dsh-tool-call-timeout-policy`，位于新的 `packages/guard/` 组；cordis.yml 的 `id` 仍可为 `timeout-policy`。
|
||||
|
||||
**仅保留逐工具的超时处理。** 这是 `bash` 和 `web_fetch` 的既有形态，也与 Claude Code 和 Codex 对 shell 命令的做法一致。它对 web 类工具不利，因为每个新的支持超时的工具都必须自行选择校验方式、上限语义、文档、快照和分类。插件集中了策略和分类，让每个工具的 schema 专注于业务输入。
|
||||
|
||||
**立即将所有超时策略移出 bash-local。** 长期来看更干净——bash-local 将成为纯子进程执行器，所有调用方自行管理截止时间。但作为第一步不合适，因为钩子直接调用 `ctx.bash`，且 bash 模型工具的前台/后台语义与工具调用生命周期不同。保留 `BASH_TIMEOUT` 维持了这些路径的稳定，同时让工具调用超时在更简单的工具上先行验证。
|
||||
**立即将所有超时策略移出 bash-local。** 长期来看更干净——bash-local 将成为纯子进程执行器，所有调用方自行管理截止时间。但作为第一步不合适，因为钩子直接调用 `ctx.shell`，且 bash 模型工具的前台/后台语义与工具调用生命周期不同。保留 `BASH_TIMEOUT` 维持了这些路径的稳定，同时让工具调用超时在更简单的工具上先行验证。
|
||||
|
||||
**为所有工具使用全局默认预算。** 方便，但会让工具作者意外：任何偶然运行超过全局预算的工具在插件加载后就会开始失败。逐工具声明预算使采纳成为有意的行为。
|
||||
|
||||
|
||||
@@ -2,5 +2,5 @@
|
||||
# side as of the last confirmed-consistent state. Both languages carry equal authority;
|
||||
# after editing either side, bring the other along and re-record with:
|
||||
# pnpm run verify-translation-pairing --write .agents/notes/implemented/architecture/2026-07-08-tool-output-spill-files.md
|
||||
2026-07-08-tool-output-spill-files.md: 3c24bc9ed754726b70e833627a22167c8256162c
|
||||
2026-07-08-tool-output-spill-files.zh.md: 777958a97b538c29fb58ffec0db86deb7e2202d7
|
||||
2026-07-08-tool-output-spill-files.md: 4d1d4b7b665f34b362df2f8c8aeb06d96bf2668f
|
||||
2026-07-08-tool-output-spill-files.zh.md: 1a3387830861c6d05318024870325523b8eeee36
|
||||
|
||||
@@ -113,8 +113,8 @@ ctx.tools.register(defineTool({
|
||||
With `dsh-spill-policy` configured, a large formatted fetch result is automatically retained and spilled. A deployment demonstrates the behavior by setting the provider resource cap higher than the policy cap:
|
||||
|
||||
```yaml
|
||||
- id: web-fetch-local
|
||||
name: '@deepseek-ai/dsh-web-fetch-local'
|
||||
- id: web-fetch-http
|
||||
name: '@deepseek-ai/dsh-web-fetch-http'
|
||||
config:
|
||||
maxBodyChars: 500000
|
||||
|
||||
@@ -127,13 +127,13 @@ With `dsh-spill-policy` configured, a large formatted fetch result is automatica
|
||||
maxInlineBytes: 50000
|
||||
```
|
||||
|
||||
This separation is important. `web-fetch-local` still owns resource caps (`maxResponseBytes`, `maxBodyChars`) to protect network, memory, and decoding work. `spill-policy` owns only the model-facing context cap after the result already exists. If the provider already returned `truncated: true`, the spill file contains the full formatted result the tool returned, not the full original webpage; the policy does not claim otherwise.
|
||||
This separation is important. `web-fetch-http` still owns resource caps (`maxResponseBytes`, `maxBodyChars`) to protect network, memory, and decoding work. `spill-policy` owns only the model-facing context cap after the result already exists. If the provider already returned `truncated: true`, the spill file contains the full formatted result the tool returned, not the full original webpage; the policy does not claim otherwise.
|
||||
|
||||
## Relationship to retention and early spill
|
||||
|
||||
Retention is separate from spill storage:
|
||||
|
||||
- `@deepseek-ai/dsh-retention` owns preview mechanics (`TextRetainer`, `ItemRetainer`, and omitted metadata).
|
||||
- `@deepseek-ai/dsh-output-retention` owns preview mechanics (`TextRetainer`, `ItemRetainer`, and omitted metadata).
|
||||
- `@deepseek-ai/dsh-spill` owns saving final text and returning a locator plus retrieval hint.
|
||||
- `@deepseek-ai/dsh-spill-policy` applies the default final-result policy in the tool pipeline, composing the two.
|
||||
|
||||
@@ -151,7 +151,7 @@ Those cases can consume `ctx.spillStore` directly in later work. They are not pa
|
||||
- No per-tool retention configuration in v1.
|
||||
- No model-facing timeout/truncation arguments.
|
||||
- No migration of `read` output into spill files.
|
||||
- No replacement for provider/resource caps such as `web-fetch-local.maxBodyChars`.
|
||||
- No replacement for provider/resource caps such as `web-fetch-http.maxBodyChars`.
|
||||
- No bash temp-file normalization or subagent rollout capture in the first cut.
|
||||
|
||||
## Deferred
|
||||
@@ -190,6 +190,6 @@ The policy can become too large if it starts owning tool-specific semantics. It
|
||||
|
||||
**Use `ctx.fs.writeText` or the model-facing `write` tool.** Rejected: workspace filesystem writes carry project-file semantics, write/edit policy, observation state, and user-facing side effects. Spill files are runtime artifacts, not model-authored workspace edits. The existing `read` tool may inspect them later, but creation belongs to the runtime spill seam.
|
||||
|
||||
**Let `web-fetch-local` fetch without caps and rely on spill-policy.** Rejected: spill-policy runs after the final tool result exists and cannot protect network, memory, or decoding resources. Provider resource caps stay mandatory.
|
||||
**Let `web-fetch-http` fetch without caps and rely on spill-policy.** Rejected: spill-policy runs after the final tool result exists and cannot protect network, memory, or decoding resources. Provider resource caps stay mandatory.
|
||||
|
||||
**Merge retention into spill.** Rejected: retention and spill have different responsibilities. `TextRetainer`/`ItemRetainer` decide what preview is kept and what was omitted; spill storage only saves the final text the policy asks it to save.
|
||||
|
||||
@@ -113,8 +113,8 @@ ctx.tools.register(defineTool({
|
||||
配置 `dsh-spill-policy` 后，格式化后的大型 fetch 结果会自动保留并 spill。部署通过把提供方资源上限设得高于策略上限来展示此行为：
|
||||
|
||||
```yaml
|
||||
- id: web-fetch-local
|
||||
name: '@deepseek-ai/dsh-web-fetch-local'
|
||||
- id: web-fetch-http
|
||||
name: '@deepseek-ai/dsh-web-fetch-http'
|
||||
config:
|
||||
maxBodyChars: 500000
|
||||
|
||||
@@ -127,13 +127,13 @@ ctx.tools.register(defineTool({
|
||||
maxInlineBytes: 50000
|
||||
```
|
||||
|
||||
这项分离很重要。`web-fetch-local` 仍负责资源上限（`maxResponseBytes`、`maxBodyChars`），用来保护网络、内存和解码工作。`spill-policy` 只负责结果已经存在后针对模型上下文的上限。如果提供方已经返回 `truncated: true`，spill 文件包含的是工具返回的完整格式化结果，而不是原始网页全文；策略不会做出其他承诺。
|
||||
这项分离很重要。`web-fetch-http` 仍负责资源上限（`maxResponseBytes`、`maxBodyChars`），用来保护网络、内存和解码工作。`spill-policy` 只负责结果已经存在后针对模型上下文的上限。如果提供方已经返回 `truncated: true`，spill 文件包含的是工具返回的完整格式化结果，而不是原始网页全文；策略不会做出其他承诺。
|
||||
|
||||
## 与保留和提前 spill 的关系
|
||||
|
||||
保留与 spill 存储相互独立：
|
||||
|
||||
- `@deepseek-ai/dsh-retention` 负责预览机制（`TextRetainer`、`ItemRetainer` 和省略元数据）。
|
||||
- `@deepseek-ai/dsh-output-retention` 负责预览机制（`TextRetainer`、`ItemRetainer` 和省略元数据）。
|
||||
- `@deepseek-ai/dsh-spill` 负责保存最终文本，并返回定位符与检索提示。
|
||||
- `@deepseek-ai/dsh-spill-policy` 在工具流水线中应用默认的最终结果策略，将前两者组合起来。
|
||||
|
||||
@@ -151,7 +151,7 @@ ctx.tools.register(defineTool({
|
||||
- v1 不增加逐工具的保留配置。
|
||||
- 不增加面向模型的超时／截断参数。
|
||||
- 不把 `read` 输出迁移到 spill 文件。
|
||||
- 不取代 `web-fetch-local.maxBodyChars` 等提供方／资源上限。
|
||||
- 不取代 `web-fetch-http.maxBodyChars` 等提供方／资源上限。
|
||||
- 第一版不统一 bash 临时文件，也不采集 subagent 执行轨迹。
|
||||
|
||||
## 延后事项
|
||||
@@ -190,6 +190,6 @@ ctx.tools.register(defineTool({
|
||||
|
||||
**使用 `ctx.fs.writeText` 或面向模型的 `write` 工具。** 不予采纳：工作区文件系统写入带有项目文件语义、写入／编辑策略、观察状态和面向用户的副作用。spill 文件是运行时产物，不是由模型编写的工作区改动。现有 `read` 工具之后可以检查它们，但创建操作属于运行时 spill seam。
|
||||
|
||||
**让 `web-fetch-local` 不受限地抓取，只依靠 spill-policy。** 不予采纳：spill-policy 在最终工具结果已经存在之后才运行，无法保护网络、内存或解码资源。提供方资源上限仍然必须存在。
|
||||
**让 `web-fetch-http` 不受限地抓取，只依靠 spill-policy。** 不予采纳：spill-policy 在最终工具结果已经存在之后才运行，无法保护网络、内存或解码资源。提供方资源上限仍然必须存在。
|
||||
|
||||
**把保留合并进 spill 机制。** 不予采纳：保留与 spill 职责不同。`TextRetainer`／`ItemRetainer` 决定保留哪部分预览、又省略了什么；spill 存储只负责保存策略要求的最终文本。
|
||||
|
||||
@@ -2,5 +2,5 @@
|
||||
# side as of the last confirmed-consistent state. Both languages carry equal authority;
|
||||
# after editing either side, bring the other along and re-record with:
|
||||
# pnpm run verify-translation-pairing --write .agents/notes/implemented/architecture/2026-07-10-after-call-compaction-pressure-and-overflow-recovery.md
|
||||
2026-07-10-after-call-compaction-pressure-and-overflow-recovery.md: 08715d3ad6ad7a2346327861e7134b10da184f00
|
||||
2026-07-10-after-call-compaction-pressure-and-overflow-recovery.zh.md: 6dd4aeff6067874b403f6dfc08fd42d4e5963a91
|
||||
2026-07-10-after-call-compaction-pressure-and-overflow-recovery.md: 89c8917484c11057ff2e6b55e2311e83f724f7ed
|
||||
2026-07-10-after-call-compaction-pressure-and-overflow-recovery.zh.md: 96f7aa1b0ea4e44ca6318caca2c785d898b1d00d
|
||||
|
||||
@@ -18,27 +18,27 @@ Successful calls are not the only pressure signal. A provider can reject a reque
|
||||
|
||||
Compact-basic wraps `agent/pre-step` before each proposed request. At a continuation boundary the preceding assistant output, every dispatched or synthetic tool result, post-tool context, and steering are already durable, so pressure policy sees the complete successful-call state without splitting an assistant tool call from its result. At the initial boundary a headerless session has no completed routed request and produces no pressure work. Compact-basic contains operational failures, warns, and delegates without rejecting the proposed step.
|
||||
|
||||
`dsh-compact-basic` reads the exact latest routed model from the durable request header only to establish that a completed route exists, then asks the singleton `ctx.tokenMeter` to measure the canonical logged envelope and current surface. It does not fall back to `AgentOptions.model` for automatic pressure. A headerless session has no completed routed request to assess and produces no work; any durable non-empty model name uses the same estimator. Operational measurement or summarization failures warn and continue from the latest durable surface: full history before any replacement, or the pruned surface if pruning already landed.
|
||||
`dsh-compaction-basic` reads the exact latest routed model from the durable request header only to establish that a completed route exists, then asks the singleton `ctx.tokenMeter` to measure the canonical logged envelope and current surface. It does not fall back to `AgentOptions.model` for automatic pressure. A headerless session has no completed routed request to assess and produces no work; any durable non-empty model name uses the same estimator. Operational measurement or summarization failures warn and continue from the latest durable surface: full history before any replacement, or the pruned surface if pruning already landed.
|
||||
|
||||
### Request recovery is limited to the final model boundary
|
||||
|
||||
`agent/request-error` represents terminal failures from the final adapter boundary. Adapter selection, dispatch, iterator construction, and iteration throws become terminal `error` or `aborted` finishes before the agent loop consumes them; adapter-emitted terminal finishes enter the same path. Prompt assembly, request middleware, request logging, result processing, tools, step listeners, and cleanup remain ordinary failures. [Terminal LLM stream failures](2026-07-29-terminal-llm-stream-failures.md) owns this normalization boundary.
|
||||
|
||||
The failed step closes before recovery runs. A handling listener repairs durable state, returns `{ kind: 'retry' }`, and stops waterfall delegation. The loop then closes the failed turn and opens one retry turn from the durable log without an intervening idle notification. Retry policy and attempt counts remain plugin-owned; compact-basic clears its per-agent overflow count when the chain reaches terminal `agent/settled`. Both DeepSeek adapters normalize recognized provider context-limit failures to `CONTEXT_WINDOW_EXCEEDED`. The [retry-action decision](../simplification/2026-07-27-request-error-retry-action.md) owns the return boundary.
|
||||
The failed step closes before recovery runs. A handling listener repairs durable state, returns `{ kind: 'retry' }`, and stops waterfall delegation. The loop then closes the failed turn and opens one retry turn from the durable log without an intervening idle notification. Retry policy and attempt counts remain plugin-owned; compaction-basic clears its per-agent overflow count when the chain reaches terminal `agent/settled`. Both DeepSeek adapters normalize recognized provider context-limit failures to `CONTEXT_WINDOW_EXCEEDED`. The [retry-action decision](../simplification/2026-07-27-request-error-retry-action.md) owns the return boundary.
|
||||
|
||||
If cancellation lands after assistant tool calls are durable but before all calls dispatch, the loop records a synthetic `tool/call` and aborted `tool/result` pair for every undispatched call before following the normal abort path. The surface therefore never retains orphaned durable tool calls merely because cancellation won the race.
|
||||
|
||||
### CompactService exposes intent, not token accounting
|
||||
### CompactionEngine exposes intent, not token accounting
|
||||
|
||||
`CompactService.compactIfNeeded(agent, trigger, signal)` accepts `trigger: 'pressure' | 'context-overflow'`. The interface gains no estimation methods or token types; `ctx.tokenMeter` remains the reusable accounting owner.
|
||||
`CompactionEngine.compactIfNeeded(agent, trigger, signal)` accepts `trigger: 'pressure' | 'context-overflow'`. The interface gains no estimation methods or token types; `ctx.tokenMeter` remains the reusable accounting owner.
|
||||
|
||||
For `pressure`, compact-basic resolves the durable provider/model target's adapter-owned capacity and exact-target policy, then applies the resulting threshold and retained-tail budgets to one unified `ctx.tokenMeter.measure()` result. Below pressure it returns without pruning. Once pressure qualifies, optional `ctx.toolResultPrune` rewrites oversized current results and compact-basic remeasures through the same meter; safe pressure skips the model call, while remaining pressure selects and summarizes from the pruned surface. The same singleton meter owns range pricing, cited source-event accounting, shadowed token counts, and non-shrinking-summary rejection. Common defaults remain threshold ratio `0.8`, retained-history ratio `0.16`, summarization provider/model `''`, `maxTokens: 8192`, `compactionRetries: 1`, and `auto: true`; optional `modelPolicies` entries override them for an exact provider/model pair.
|
||||
For `pressure`, compaction-basic resolves the durable provider/model target's adapter-owned capacity and exact-target policy, then applies the resulting threshold and retained-tail budgets to one unified `ctx.tokenMeter.measure()` result. Below pressure it returns without pruning. Once pressure qualifies, optional `ctx.toolResultPruner` rewrites oversized current results and compaction-basic remeasures through the same meter; safe pressure skips the model call, while remaining pressure selects and summarizes from the pruned surface. The same singleton meter owns range pricing, cited source-event accounting, shadowed token counts, and non-shrinking-summary rejection. Common defaults remain threshold ratio `0.8`, retained-history ratio `0.16`, summarization provider/model `''`, `maxTokens: 8192`, `compactionRetries: 1`, and `auto: true`; optional `modelPolicies` entries override them for an exact provider/model pair.
|
||||
|
||||
For canonical overflow, compact-basic requires no capacity metadata and bypasses scalar pressure and the normal retained-token budget. It prunes first, then chooses the maximal tool-balanced head range while leaving the newest indivisible unit and attempts one shrinking summary compaction under the same signal when a range exists. The automatic listener snapshots `session.surface.replaceGeneration` and returns `{ kind: 'retry' }` whenever pruning or summarization increases it. This remains true when pruning lands before later summary work throws; cancellation still wins. A backend returning a result without replacement cannot authorize retry, while pruning-only progress can authorize a retry without a `CompactionResult`.
|
||||
For canonical overflow, compaction-basic requires no capacity metadata and bypasses scalar pressure and the normal retained-token budget. It prunes first, then chooses the maximal tool-balanced head range while leaving the newest indivisible unit and attempts one shrinking summary compaction under the same signal when a range exists. The automatic listener snapshots `session.surface.replaceGeneration` and returns `{ kind: 'retry' }` whenever pruning or summarization increases it. This remains true when pruning lands before later summary work throws; cancellation still wins. A backend returning a result without replacement cannot authorize retry, while pruning-only progress can authorize a retry without a `CompactionResult`.
|
||||
|
||||
`maxOverflowRetries` is optional and defaults to `1`; `0` disables overflow recovery without disabling pressure. `auto: false` registers neither automatic listener. Noncanonical errors, exhausted attempts, an already-aborted signal, a missing routed model, no safe range, no generation change, and recovery throws before any replacement all delegate to the next listener. With no later recovery, the loop reports the original provider error object and code. A recovery throw after generation advances authorizes retry from durable progress; cancellation or disposal remains authoritative even if recovery work completes concurrently.
|
||||
|
||||
The default summarizer resolves explicit configuration, then the latest logged route, then agent options. Because direct `llm/stream` middleware may reroute that auxiliary call, `compact/summary.{provider, model}` records the final mutable `GenerateOptions` target observed after dispatch rather than the pre-waterfall candidate.
|
||||
The default summarizer resolves explicit configuration, then the latest logged route, then agent options. Because direct `llm/stream` middleware may reroute that auxiliary call, `compaction/summary.{provider, model}` records the final mutable `GenerateOptions` target observed after dispatch rather than the pre-waterfall candidate.
|
||||
|
||||
## Testing
|
||||
|
||||
@@ -49,7 +49,7 @@ Unit tests cover the final-adapter normalization boundary, closed-turn retry num
|
||||
- **Add compaction-only fields to pre-step** — rejected because the canonical durable session and token meter already own the measurement input; the generic lifecycle need not carry a second envelope.
|
||||
- **Retry the same numbered step** — rejected because recovery appends durable events after the failed boundary. A new step preserves balanced nesting and reconstructability.
|
||||
- **Retry whenever `compactIfNeeded` returns a result** — rejected because a custom backend can report success without changing model-visible state. `replaceGeneration` is the authoritative proof.
|
||||
- **Let compact-basic parse provider wording** — rejected because classification belongs at adapters and must cover both thrown and in-band delivery.
|
||||
- **Let compaction-basic parse provider wording** — rejected because classification belongs at adapters and must cover both thrown and in-band delivery.
|
||||
- **Fall back to `AgentOptions.model` when no durable route exists** — rejected because automatic policy must describe a completed logged request. Headerless pressure and recovery delegate unchanged.
|
||||
|
||||
## Consequences
|
||||
|
||||
@@ -18,27 +18,27 @@ Status: implemented
|
||||
|
||||
Compact-basic 会在每个拟议请求之前包装 `agent/pre-step`。在续步边界，前一条 assistant 输出、所有已分发或合成的工具结果、工具后上下文与 steering 都已经持久化，因此压力策略能看到完整的成功调用状态，同时不会拆开 assistant 工具调用与其结果。初始边界上的无 header 会话尚无已完成路由请求，因此不执行压力工作。Compact-basic 会在内部处理操作性失败、发出警告并继续委托，不会 reject 拟议步骤。
|
||||
|
||||
`dsh-compact-basic` 从持久请求头读取精确的最新实际路由模型，只用它确认已经存在已完成的路由，随后让单例 `ctx.tokenMeter` 计量规范日志信封与当前表层。自动压力不会回退到 `AgentOptions.model`。没有请求头的会话尚无已完成路由请求可供判断，因此不执行工作；任意持久记录的非空模型名都使用同一个估算器。操作性的计量或摘要失败会发出警告，并从最新持久表层继续：任何替换发生前使用完整历史；若剪枝已经落盘，则使用已剪枝表层。
|
||||
`dsh-compaction-basic` 从持久请求头读取精确的最新实际路由模型，只用它确认已经存在已完成的路由，随后让单例 `ctx.tokenMeter` 计量规范日志信封与当前表层。自动压力不会回退到 `AgentOptions.model`。没有请求头的会话尚无已完成路由请求可供判断，因此不执行工作；任意持久记录的非空模型名都使用同一个估算器。操作性的计量或摘要失败会发出警告，并从最新持久表层继续：任何替换发生前使用完整历史；若剪枝已经落盘，则使用已剪枝表层。
|
||||
|
||||
### 请求恢复只覆盖最终模型边界
|
||||
|
||||
`agent/request-error` 表示来自最终适配器边界的终止失败。适配器选择、分发、iterator 构造与迭代抛出会在 agent loop（智能体循环）消费前成为终止 `error` 或 `aborted` finish；适配器直接发出的终止 finish 进入同一路径。提示词装配、请求 middleware、请求日志、结果处理、工具、step 监听器与清理仍属于普通失败。[LLM（大语言模型）流的终止失败](2026-07-29-terminal-llm-stream-failures.md)规定这一规范化边界。
|
||||
|
||||
恢复运行前，失败 step 已经关闭。负责处理的监听器修复持久状态、返回 `{ kind: 'retry' }`，并停止 waterfall（瀑布式事件）委托。循环随后关闭失败 turn，并从持久日志开启一个重试 turn，中间不发布空闲通知。重试策略与尝试计数由插件自己拥有；compact-basic 在链路到达终态 `agent/settled` 时清除对应 agent 的溢出计数。两个 DeepSeek 适配器都把识别出的提供方上下文限制错误规范化为 `CONTEXT_WINDOW_EXCEEDED`。[重试动作决策](../simplification/2026-07-27-request-error-retry-action.md)规定这一返回边界。
|
||||
恢复运行前，失败 step 已经关闭。负责处理的监听器修复持久状态、返回 `{ kind: 'retry' }`，并停止 waterfall（瀑布式事件）委托。循环随后关闭失败 turn，并从持久日志开启一个重试 turn，中间不发布空闲通知。重试策略与尝试计数由插件自己拥有；compaction-basic 在链路到达终态 `agent/settled` 时清除对应 agent 的溢出计数。两个 DeepSeek 适配器都把识别出的提供方上下文限制错误规范化为 `CONTEXT_WINDOW_EXCEEDED`。[重试动作决策](../simplification/2026-07-27-request-error-retry-action.md)规定这一返回边界。
|
||||
|
||||
如果取消发生在 assistant 工具调用已经持久化之后、所有调用完成分发之前，循环会为每个尚未分发的调用记录一对合成的 `tool/call` 与 aborted `tool/result`，随后进入正常中止路径。因此，表层不会仅因取消赢得竞态而留下孤立的持久工具调用。
|
||||
|
||||
### CompactService 暴露意图，而不拥有 token 核算
|
||||
### CompactionEngine 暴露意图，而不拥有 token 核算
|
||||
|
||||
`CompactService.compactIfNeeded(agent, trigger, signal)` 接收 `trigger: 'pressure' | 'context-overflow'`。接口不增加估算方法或 token 类型；`ctx.tokenMeter` 继续作为可复用的核算所有者。
|
||||
`CompactionEngine.compactIfNeeded(agent, trigger, signal)` 接收 `trigger: 'pressure' | 'context-overflow'`。接口不增加估算方法或 token 类型；`ctx.tokenMeter` 继续作为可复用的核算所有者。
|
||||
|
||||
对于 `pressure`，compact-basic 先解析持久提供方/模型目标对应适配器所维护的容量与精确目标策略，再把得到的阈值与保留尾部预算应用到一次统一的 `ctx.tokenMeter.measure()` 结果。未达到压力阈值时直接返回，不执行剪枝。压力达到条件后，可选的 `ctx.toolResultPrune` 会改写当前表层中过大的工具结果，compact-basic 再通过同一个 meter 重新计量；若压力已降至安全水平则跳过模型调用，否则从已剪枝表层选择范围并生成摘要。范围定价、引用的源事件计量、被遮蔽 token 数与非缩小摘要拒绝也由同一个单例 meter 完成。通用默认值保持为阈值比例 `0.8`、保留历史比例 `0.16`、摘要提供方/模型 `''`、`maxTokens: 8192`、`compactionRetries: 1` 与 `auto: true`；可选 `modelPolicies` 项可以按精确提供方/模型组合覆盖这些值。
|
||||
对于 `pressure`，compaction-basic 先解析持久提供方/模型目标对应适配器所维护的容量与精确目标策略，再把得到的阈值与保留尾部预算应用到一次统一的 `ctx.tokenMeter.measure()` 结果。未达到压力阈值时直接返回，不执行剪枝。压力达到条件后，可选的 `ctx.toolResultPruner` 会改写当前表层中过大的工具结果，compaction-basic 再通过同一个 meter 重新计量；若压力已降至安全水平则跳过模型调用，否则从已剪枝表层选择范围并生成摘要。范围定价、引用的源事件计量、被遮蔽 token 数与非缩小摘要拒绝也由同一个单例 meter 完成。通用默认值保持为阈值比例 `0.8`、保留历史比例 `0.16`、摘要提供方/模型 `''`、`maxTokens: 8192`、`compactionRetries: 1` 与 `auto: true`；可选 `modelPolicies` 项可以按精确提供方/模型组合覆盖这些值。
|
||||
|
||||
对于规范化溢出，compact-basic 不要求容量元数据，并绕过标量压力与普通保留 token 预算。它先执行剪枝，再在保留最新不可分割单元的同时选择最大的工具配对平衡头部范围；存在范围时，才在同一 signal 下尝试一次缩小摘要压缩。自动监听器先对 `session.surface.replaceGeneration` 建立快照，剪枝或摘要让 generation 增加时就返回 `{ kind: 'retry' }`。即使剪枝先落盘而后续摘要工作抛错，这条规则仍然成立；取消依然优先。后端若只返回结果但没有替换表层，不能授权重试；只有剪枝取得进展时，即使没有 `CompactionResult` 也可以授权重试。
|
||||
对于规范化溢出，compaction-basic 不要求容量元数据，并绕过标量压力与普通保留 token 预算。它先执行剪枝，再在保留最新不可分割单元的同时选择最大的工具配对平衡头部范围；存在范围时，才在同一 signal 下尝试一次缩小摘要压缩。自动监听器先对 `session.surface.replaceGeneration` 建立快照，剪枝或摘要让 generation 增加时就返回 `{ kind: 'retry' }`。即使剪枝先落盘而后续摘要工作抛错，这条规则仍然成立；取消依然优先。后端若只返回结果但没有替换表层，不能授权重试；只有剪枝取得进展时，即使没有 `CompactionResult` 也可以授权重试。
|
||||
|
||||
`maxOverflowRetries` 可选且默认为 `1`；`0` 只禁用溢出恢复，不会禁用压力检查。`auto: false` 不注册任何自动监听器。非规范化错误、尝试耗尽、已经中止的 signal、缺失路由模型、没有安全范围、generation 未变化，以及在任何替换之前恢复抛错，都会委托给下一个监听器。若没有后续恢复，循环报告原始提供方错误对象与代码。generation 增加后的恢复抛错会基于持久进展授权重试；即使恢复工作并发完成，取消或 dispose（资源释放）仍具有最终优先级。
|
||||
|
||||
默认摘要器依次解析显式配置、最近记录的路由与 agent options。因为直接 `llm/stream` 中间件可以重新路由该辅助调用，`compact/summary.{provider, model}` 记录分发后观察到的可变 `GenerateOptions` 最终目标，而不是 waterfall 之前的候选值。
|
||||
默认摘要器依次解析显式配置、最近记录的路由与 agent options。因为直接 `llm/stream` 中间件可以重新路由该辅助调用，`compaction/summary.{provider, model}` 记录分发后观察到的可变 `GenerateOptions` 最终目标，而不是 waterfall 之前的候选值。
|
||||
|
||||
## 测试
|
||||
|
||||
@@ -49,7 +49,7 @@ Compact-basic 会在每个拟议请求之前包装 `agent/pre-step`。在续步
|
||||
- **向 pre-step 增加压缩专用字段**——不予采纳，因为规范持久会话与 token meter 已拥有计量输入；通用生命周期不需要携带第二份信封。
|
||||
- **重试相同编号的 step**——不予采纳，因为恢复会在失败边界之后追加持久事件。新 step 保持平衡嵌套与可重建性。
|
||||
- **只要 `compactIfNeeded` 返回结果就重试**——不予采纳，因为自定义后端可能报告成功却没有改变模型可见状态。`replaceGeneration` 才是权威证明。
|
||||
- **让 compact-basic 解析提供方措辞**——不予采纳，因为分类属于适配器，而且必须同时覆盖抛出式与带内交付。
|
||||
- **让 compaction-basic 解析提供方措辞**——不予采纳，因为分类属于适配器，而且必须同时覆盖抛出式与带内交付。
|
||||
- **没有持久路由时回退到 `AgentOptions.model`**——不予采纳，因为自动策略必须描述已完成且已记录的请求。没有请求头的压力检查与恢复会原样委托。
|
||||
|
||||
## 后果
|
||||
|
||||
@@ -2,5 +2,5 @@
|
||||
# side as of the last confirmed-consistent state. Both languages carry equal authority;
|
||||
# after editing either side, bring the other along and re-record with:
|
||||
# pnpm run verify-translation-pairing --write .agents/notes/implemented/architecture/2026-07-10-single-file-executable-sdk-runtime-distribution.md
|
||||
2026-07-10-single-file-executable-sdk-runtime-distribution.md: e45f566ee542fdab4ae55b45ff6808169e6e7cf6
|
||||
2026-07-10-single-file-executable-sdk-runtime-distribution.zh.md: ca076b62f928bc146d548b0e2996dfd41a2bf1e3
|
||||
2026-07-10-single-file-executable-sdk-runtime-distribution.md: fa2f86893b730aa1ba020bd568d268ec8d9d6239
|
||||
2026-07-10-single-file-executable-sdk-runtime-distribution.zh.md: 509bec18edb9923dd4d60d4ecf30d4fbcd9cc6d5
|
||||
|
||||
@@ -6,7 +6,7 @@ English | [中文](2026-07-10-single-file-executable-sdk-runtime-distribution.zh
|
||||
|
||||
## Problem
|
||||
|
||||
DeepSeek Harness needs a dedicated SDK distribution form for the Python library — no Node installation, runs directly on the target platform: a single-file executable (hereafter "the exe") that exposes a stdio JSON-RPC serving interface (`HarnessSdkServer`, the Python SDK's peer), where the plugins and configuration actually booted are decided entirely by a `cordis.yml` supplied from outside the exe.
|
||||
DeepSeek Harness needs a dedicated SDK distribution form for the Python library — no Node installation, runs directly on the target platform: a single-file executable (hereafter "the exe") that exposes a stdio JSON-RPC serving interface (`HarnessSdkJsonRpcServer`, the Python SDK's peer), where the plugins and configuration actually booted are decided entirely by a `cordis.yml` supplied from outside the exe.
|
||||
|
||||
- The JSONRPC protocol for talking to the Python SDK is already validated
|
||||
- A standardized way for cordis.yml to load every plugin (ESModule) is needed
|
||||
@@ -27,8 +27,8 @@ Terminology reminder: pkg's `/snapshot` VFS has nothing to do with this repo's t
|
||||
|
||||
The deterministic protocol implementation (`server.ts` / `transport.ts`) lands as two packages on the existing `acp/acp` + `examples/acp-demo` pattern — the serving surface is itself a plugin:
|
||||
|
||||
- [`packages/sdk/server`](../../../../packages/sdk/server/README.md) (`@deepseek-ai/dsh-jsonrpc`): the pure protocol plugin; on apply it mounts `HarnessSdkServer` plus a line-delimited JSON-RPC transport on the process stdio, with disposal through `ctx.effect()`. Whether to serve is decided by `cordis.yml`; a yml that does not mount it is a legitimate process that does not serve. Protocol-level exit belongs to the plugin (after answering and flushing the `shutdown` response it disposes the root runtime so persistence drains, then `exit(0)`; an HMR-style unload only stops the service without exiting the process).
|
||||
- [`packages/examples/jsonrpc-demo`](../../../../packages/examples/jsonrpc-demo/README.md) (`@deepseek-ai/dsh-jsonrpc-demo`): a thin app bin — `installFailLoud` + `loadEnv` + config discovery + `boot()` from [`dsh-app-boot`](../../../../packages/boot/app-boot/src/index.ts), done once boot completes; the server is brought up by the `dsh-jsonrpc` entry in the yml. Its only dependency is app-boot. Process-level exit belongs to the bin (stdin EOF/SIGTERM → dispose then 0, SIGINT → 130).
|
||||
- [`packages/sdk/server`](../../../../packages/sdk/server/README.md) (`@deepseek-ai/dsh-sdk-jsonrpc-server`): the pure protocol plugin; on apply it mounts `HarnessSdkJsonRpcServer` plus a line-delimited JSON-RPC transport on the process stdio, with disposal through `ctx.effect()`. Whether to serve is decided by `cordis.yml`; a yml that does not mount it is a legitimate process that does not serve. Protocol-level exit belongs to the plugin (after answering and flushing the `shutdown` response it disposes the root runtime so persistence drains, then `exit(0)`; an HMR-style unload only stops the service without exiting the process).
|
||||
- [`packages/examples/jsonrpc-demo`](../../../../packages/examples/jsonrpc-demo/README.md) (`@deepseek-ai/dsh-sdk-jsonrpc-demo`): a thin app bin — `installFailLoud` + `loadEnv` + config discovery + `boot()` from [`dsh-app-boot`](../../../../packages/boot/app-boot/src/index.ts), done once boot completes; the server is brought up by the `dsh-sdk-jsonrpc-server` entry in the yml. Its only dependency is app-boot. Process-level exit belongs to the bin (stdin EOF/SIGTERM → dispose then 0, SIGINT → 130).
|
||||
|
||||
Config discovery has two channels and fails loudly when both are missing: the `DSH_CORDIS_CONFIG` environment variable first (the SDK client convention), then an argv positional argument; no default path and no built-in fallback whatsoever — "the plugins actually booted are decided by an external cordis.yml" is a hard semantic.
|
||||
|
||||
@@ -40,25 +40,25 @@ The deploy root is [`python/sdk-runtime/package.json`](../../../../python/sdk-ru
|
||||
|
||||
### Build pipeline and artifacts
|
||||
|
||||
[`scripts/build-exe-for-python-sdk.ts`](../../../../scripts/build-exe-for-python-sdk.ts): runtime closure verification → `pnpm run build` → (after clearing) `pnpm --filter dsh-jsonrpc-agent-pkg deploy --legacy --prod --config.node-linker=hoisted --config.auto-install-peers=false --config.link-workspace-packages=true` **directly into** `python/sdk-runtime/src/deepseek_harness_runtime/runtime/node/` → restore any direct workspace package that legacy deploy hoisted back under the source manifest's `node_modules`, omitting its package-local dependency tree and rejecting any remaining manifest gap → replace every staged dependency symlink with its target bytes, remove package-manager `.bin` links, and fail if any symlink remains → inject the pkg configuration (`bin` points at `node_modules/@deepseek-ai/dsh-jsonrpc-demo/lib/packaged-bin.js` inside the closure, `assets` is a full glob — dynamic import is invisible to pkg's static analysis, so everything must be packed in explicitly) → stage the target `node-pty` addon → one `pkg --sea` per target → the executables `dsh-jsonrpc-agent-pkg-<platform>-<arch>` land in `dist-exe/` and are copied back into the runtime directory. Linux installs build `pty.node` from source; CI rebuilds that addon inside the matching manylinux 2.28 container before packaging, and the builder copies it from the root install into the staged closure because legacy deploy omits that side-effect directory. macOS uses its target prebuild and emits the required `-spawn-helper` beside the executable. CI treats these products as intermediate test inputs and retains their platform wheels. All four deploy flags are grounded in measurement: `--legacy` is the mandatory path with inject-workspace-packages off; hoisted gives pkg a stable single-instance layout that the explicit materialization pass makes symlink-free; disabling automatic peer installation prevents undeclared peers from expanding the closure; link-workspace-packages selects direct workspace dependencies. [`pnpm-workspace.yaml`](../../../../pnpm-workspace.yaml) overrides the transitive `@deepseek-ai/cosmokit` and `@deepseek-ai/schemastery` semver requests to the pinned vendor sources so legacy deploy never resolves those unpublished names from a registry.
|
||||
[`scripts/build-exe-for-python-sdk.ts`](../../../../scripts/build-exe-for-python-sdk.ts): runtime closure verification → `pnpm run build` → (after clearing) `pnpm --filter dsh-jsonrpc-agent-pkg deploy --legacy --prod --config.node-linker=hoisted --config.auto-install-peers=false --config.link-workspace-packages=true` **directly into** `python/sdk-runtime/src/deepseek_harness_runtime/runtime/node/` → restore any direct workspace package that legacy deploy hoisted back under the source manifest's `node_modules`, omitting its package-local dependency tree and rejecting any remaining manifest gap → replace every staged dependency symlink with its target bytes, remove package-manager `.bin` links, and fail if any symlink remains → inject the pkg configuration (`bin` points at `node_modules/@deepseek-ai/dsh-sdk-jsonrpc-demo/lib/packaged-bin.js` inside the closure, `assets` is a full glob — dynamic import is invisible to pkg's static analysis, so everything must be packed in explicitly) → stage the target `node-pty` addon → one `pkg --sea` per target → the executables `dsh-jsonrpc-agent-pkg-<platform>-<arch>` land in `dist-exe/` and are copied back into the runtime directory. Linux installs build `pty.node` from source; CI rebuilds that addon inside the matching manylinux 2.28 container before packaging, and the builder copies it from the root install into the staged closure because legacy deploy omits that side-effect directory. macOS uses its target prebuild and emits the required `-spawn-helper` beside the executable. CI treats these products as intermediate test inputs and retains their platform wheels. All four deploy flags are grounded in measurement: `--legacy` is the mandatory path with inject-workspace-packages off; hoisted gives pkg a stable single-instance layout that the explicit materialization pass makes symlink-free; disabling automatic peer installation prevents undeclared peers from expanding the closure; link-workspace-packages selects direct workspace dependencies. [`pnpm-workspace.yaml`](../../../../pnpm-workspace.yaml) overrides the transitive `@deepseek-ai/cosmokit` and `@deepseek-ai/schemastery` semver requests to the pinned vendor sources so legacy deploy never resolves those unpublished names from a registry.
|
||||
|
||||
CI: [`.github/workflows/build-exe-for-python-sdk.yml`](../../../../.github/workflows/build-exe-for-python-sdk.yml), called for linux-x64 by the [required Python runtime pull-request validation](../testing/2026-08-12-required-python-runtime-pull-request-ci.md), triggered explicitly by `workflow_dispatch` or the `build-exe` label for selected targets, and called for all targets by the [public publication workflow](../process/2026-08-11-python-publication-workflow.md). Native builds run on linux-x64 / linux-arm64 (`ubuntu-24.04-arm`) / macos-arm64, with `~/.pkg-cache` cached, and pkg handles macOS ad-hoc signing. Each leg drives a mock SSE model through the SDK with the default config and a custom `cordis.yml`, drives the exe directly over NDJSON JSON-RPC, verifies the JSONL and final response, and installs release-shaped wheels into a clean venv without `runtime_bin`; Linux additionally inspects both the executable and native addon's GLIBC requirements and runs in a manylinux 2.28 container, while macOS verifies that the executable's deployment target fits the wheel tag. A full three-target run retains four artifacts, each containing one release file: the platform-independent SDK wheel and three native runtime wheels; a subset dispatch retains the SDK wheel and selected runtime wheels. Bare executables and source bundles remain intermediate test inputs. [`.gitlab-ci.yml`](../../../../.gitlab-ci.yml) accepts `python-v<repository-version>` tag pipelines whose version matches the root `package.json`, builds one SDK wheel and three native runtime wheels, then a single serialized job checks and publishes all four to the project PyPI registry. Windows is a non-goal.
|
||||
|
||||
### Python SDK distribution: two carriers, exe for production, node for development
|
||||
|
||||
The Python SDK lives at [`python/`](../../../../python/README.md): `python/sdk` (the client) + `python/sdk-runtime` (the runtime carrier package). The runtime package's data directory holds the checked-in default `runtime/cordis.yml`, the build-injected platform exe and optional helper, and the build-injected `runtime/node/` closure tree. `resolve_bundled_launch_args()` automatic resolution **finds the exe only**; the node carrier is enabled only by an explicit `DSH_RUNTIME_MODE=node` (running `runtime/node/node_modules/@deepseek-ai/dsh-jsonrpc-demo/lib/packaged-bin.js`, requiring a system node ≥22.19), positioned as the development-verification channel for members of this repo, and does not enter wheel distributions.
|
||||
The Python SDK lives at [`python/`](../../../../python/README.md): `python/sdk` (the client) + `python/sdk-runtime` (the runtime carrier package). The runtime package's data directory holds the checked-in default `runtime/cordis.yml`, the build-injected platform exe and optional helper, and the build-injected `runtime/node/` closure tree. `resolve_bundled_launch_args()` automatic resolution **finds the exe only**; the node carrier is enabled only by an explicit `DSH_RUNTIME_MODE=node` (running `runtime/node/node_modules/@deepseek-ai/dsh-sdk-jsonrpc-demo/lib/packaged-bin.js`, requiring a system node ≥22.19), positioned as the development-verification channel for members of this repo, and does not enter wheel distributions.
|
||||
|
||||
[`scripts/build-python-release.py`](../../../../scripts/build-python-release.py) reads the authoritative `X.Y.Z` or prerelease version from the repository root `package.json`, converts prereleases to their PEP 440 spelling, and stages both packages at that wheel version, with `deepseek-harness-sdk` depending exactly on the matching `deepseek-harness-runtime-bin`. An optional `python-v<repository-version>` release tag is a consistency assertion and is rejected when it differs from the repository version; the source `pyproject.toml` development sentinel never determines a release version. Staging also carries the repository license into both wheels and the third-party notices into the bundled runtime wheel. The SDK is a `py3-none-any` wheel; each wheel-only runtime package contains one exe, and the macOS wheel also contains its architecture-matched helper. Runtime wheels use one of `py3-none-manylinux_2_28_x86_64`, `py3-none-manylinux_2_28_aarch64`, or the conservative `py3-none-macosx_14_0_arm64` tag for the Node 24 executable's macOS 13.5 deployment target; the Hatch hook rejects sdists, universal tags, mixed-platform payloads, missing or extra helpers, and unsupported platforms.
|
||||
|
||||
The exe's "must be explicitly configured" hard semantic is unchanged; the zero-config experience is restored by the wrapper: when the caller gave no `cordis`, named no explicit runtime, and the environment has no `DSH_CORDIS_CONFIG`, the client explicitly injects the checked-in default `cordis.yml` (agent-core + preloaded llm-deepseek + JSONL persistence + bash-local + the `dsh-jsonrpc` serving entry, with `!!js` environment-variable fallbacks) via `DSH_CORDIS_CONFIG`.
|
||||
The exe's "must be explicitly configured" hard semantic is unchanged; the zero-config experience is restored by the wrapper: when the caller gave no `cordis`, named no explicit runtime, and the environment has no `DSH_CORDIS_CONFIG`, the client explicitly injects the checked-in default `cordis.yml` (agent-core + preloaded llm-deepseek + JSONL persistence + bash-local + the `dsh-sdk-jsonrpc-server` serving entry, with `!!js` environment-variable fallbacks) via `DSH_CORDIS_CONFIG`.
|
||||
|
||||
### Naming lineage
|
||||
|
||||
`@deepseek-ai/dsh-jsonrpc-demo` (the package) → `dsh-jsonrpc-agent` (the bin) → `dsh-jsonrpc-agent-pkg` (the closure manifest; no scope prefix, deliberately sidestepping the constraints' package-shape rules for `@deepseek-ai/dsh-*`) → `dsh-jsonrpc-agent-pkg-<platform>-<arch>` (the exe artifacts). The wire `serverInfo.name` stays `deepseek-harness-sdk-runtime` (a protocol-stable value); the Python distribution names are `deepseek-harness-sdk` / `deepseek-harness-runtime-bin`, while the import modules remain `deepseek_harness` / `deepseek_harness_runtime`.
|
||||
`@deepseek-ai/dsh-sdk-jsonrpc-demo` (the package) → `dsh-jsonrpc-agent` (the bin) → `dsh-jsonrpc-agent-pkg` (the closure manifest; no scope prefix, deliberately sidestepping the constraints' package-shape rules for `@deepseek-ai/dsh-*`) → `dsh-jsonrpc-agent-pkg-<platform>-<arch>` (the exe artifacts). The wire `serverInfo.name` stays `deepseek-harness-sdk-runtime` (a protocol-stable value); the Python distribution names are `deepseek-harness-sdk` / `deepseek-harness-runtime-bin`, while the import modules remain `deepseek_harness` / `deepseek_harness_runtime`.
|
||||
|
||||
## Disposition of worker-style plugins
|
||||
|
||||
`dsh-workflow-workerthread` and `dsh-code-runtime-worker` are supported inside the exe. Their built hosts convert the sibling `lib/worker.cjs` URL with `fileURLToPath()` and pass the resulting filesystem string to `Worker`, which is the form pkg's Worker hook resolves inside the VFS. The worker entries are CommonJS because that hook compiles VFS worker files as CommonJS. The workflow engine keeps its data-URL bootstrap for unbuilt source execution; only its built sibling entry uses the filesystem string. The custom-config executable smoke loads both backends, invokes a real `run_code` call and a zero-agent `workflow` call, and requires each worker to return `42` from inside pkg's VFS.
|
||||
`dsh-workflow-worker-thread` and `dsh-code-runtime-worker-thread` are supported inside the exe. Their built hosts convert the sibling `lib/worker.cjs` URL with `fileURLToPath()` and pass the resulting filesystem string to `Worker`, which is the form pkg's Worker hook resolves inside the VFS. The worker entries are CommonJS because that hook compiles VFS worker files as CommonJS. The workflow engine keeps its data-URL bootstrap for unbuilt source execution; only its built sibling entry uses the filesystem string. The custom-config executable smoke loads both backends, invokes a real `run_code` call and a zero-agent `workflow` call, and requires each worker to return `42` from inside pkg's VFS.
|
||||
|
||||
## Testing
|
||||
|
||||
|
||||
@@ -6,7 +6,7 @@ Status: implemented
|
||||
|
||||
## 问题
|
||||
|
||||
DeepSeek Harness 需要为 Python 库专门提供一种无需安装 Node、可直接在目标平台运行的 SDK 分发形态：一个单文件可执行程序（下称 exe），通过 stdio 提供 JSON-RPC 对外服务接口（`HarnessSdkServer`，Python SDK 的对端），且实际启动的插件与配置完全由 exe 外部输入的 `cordis.yml` 决定。
|
||||
DeepSeek Harness 需要为 Python 库专门提供一种无需安装 Node、可直接在目标平台运行的 SDK 分发形态：一个单文件可执行程序（下称 exe），通过 stdio 提供 JSON-RPC 对外服务接口（`HarnessSdkJsonRpcServer`，Python SDK 的对端），且实际启动的插件与配置完全由 exe 外部输入的 `cordis.yml` 决定。
|
||||
|
||||
- 与 Python SDK 通信的 JSON-RPC 协议已经过验证
|
||||
- 需要提供一种让 `cordis.yml` 加载所有插件（ES 模块）的标准方式
|
||||
@@ -27,8 +27,8 @@ exe 使用 [@yao-pkg/pkg](https://github.com/yao-pkg/pkg)（vercel/pkg 归档后
|
||||
|
||||
确定性协议实现（`server.ts` / `transport.ts`）按 `acp/acp` + `examples/acp-demo` 的既有模式落为两包——对外服务接口本身也是插件：
|
||||
|
||||
- [`packages/sdk/server`](../../../../packages/sdk/server/README.md)（`@deepseek-ai/dsh-jsonrpc`）：纯协议插件；执行 `apply` 时，在进程 stdio 上挂载 `HarnessSdkServer` 与按行分隔的 JSON-RPC 传输层，资源释放走 `ctx.effect()`。是否提供服务由 `cordis.yml` 决定；未挂载该插件的配置会启动一个不提供此服务的合法进程。协议级退出归插件所有（应答并确保 `shutdown` 响应发送完毕后，对根运行时执行 dispose（资源释放），让待处理的持久化操作完成，再调用 `exit(0)`；HMR（热模块替换）式卸载只停止服务，不退出进程）。
|
||||
- [`packages/examples/jsonrpc-demo`](../../../../packages/examples/jsonrpc-demo/README.md)（`@deepseek-ai/dsh-jsonrpc-demo`）：轻量应用入口——`installFailLoud` + `loadEnv` + 配置发现 + [`dsh-app-boot`](../../../../packages/boot/app-boot/src/index.ts) 的 `boot()`；`boot()` 完成后入口即完成，服务器由 `cordis.yml` 中的 `dsh-jsonrpc` 条目启动。它只依赖 `app-boot`。进程级退出归 `bin` 所有（stdin EOF/SIGTERM → dispose 后返回 0，SIGINT → 130）。
|
||||
- [`packages/sdk/server`](../../../../packages/sdk/server/README.md)（`@deepseek-ai/dsh-sdk-jsonrpc-server`）：纯协议插件；执行 `apply` 时，在进程 stdio 上挂载 `HarnessSdkJsonRpcServer` 与按行分隔的 JSON-RPC 传输层，资源释放走 `ctx.effect()`。是否提供服务由 `cordis.yml` 决定；未挂载该插件的配置会启动一个不提供此服务的合法进程。协议级退出归插件所有（应答并确保 `shutdown` 响应发送完毕后，对根运行时执行 dispose（资源释放），让待处理的持久化操作完成，再调用 `exit(0)`；HMR（热模块替换）式卸载只停止服务，不退出进程）。
|
||||
- [`packages/examples/jsonrpc-demo`](../../../../packages/examples/jsonrpc-demo/README.md)（`@deepseek-ai/dsh-sdk-jsonrpc-demo`）：轻量应用入口——`installFailLoud` + `loadEnv` + 配置发现 + [`dsh-app-boot`](../../../../packages/boot/app-boot/src/index.ts) 的 `boot()`；`boot()` 完成后入口即完成，服务器由 `cordis.yml` 中的 `dsh-sdk-jsonrpc-server` 条目启动。它只依赖 `app-boot`。进程级退出归 `bin` 所有（stdin EOF/SIGTERM → dispose 后返回 0，SIGINT → 130）。
|
||||
|
||||
配置发现有两个通道，均缺失时立即报错：优先使用 `DSH_CORDIS_CONFIG` 环境变量（SDK 客户端约定），其次使用 argv 位置参数；没有默认路径或内置回退——「实际启动的插件由外部 `cordis.yml` 决定」是硬语义。
|
||||
|
||||
@@ -40,25 +40,25 @@ exe 的 VFS 内是**构建产物形态的真实包树**（各包的 `lib/` + 真
|
||||
|
||||
### 构建流水线与产物
|
||||
|
||||
[`scripts/build-exe-for-python-sdk.ts`](../../../../scripts/build-exe-for-python-sdk.ts)：运行时闭包校验 → `pnpm run build` →（清空后）`pnpm --filter dsh-jsonrpc-agent-pkg deploy --legacy --prod --config.node-linker=hoisted --config.auto-install-peers=false --config.link-workspace-packages=true` **直接写入** `python/sdk-runtime/src/deepseek_harness_runtime/runtime/node/` → 恢复被 legacy deploy 提升回源 manifest 的 `node_modules` 下的任何直接工作区包，同时省略其包内依赖树，并拒绝剩余的 manifest 缺口 → 将暂存依赖中的每个符号链接替换为目标文件内容，删除包管理器的 `.bin` 链接，并在仍有任何符号链接时失败 → 注入 pkg 配置（`bin` 指向闭包内的 `node_modules/@deepseek-ai/dsh-jsonrpc-demo/lib/packaged-bin.js`；`assets` 使用全量 glob，因为动态 `import()` 对 pkg 静态分析不可见，必须显式打入全部内容）→ 暂存目标平台的 `node-pty` addon → 每个构建目标调用一次 `pkg --sea` → 可执行文件 `dsh-jsonrpc-agent-pkg-<platform>-<arch>` 写入 `dist-exe/`，并拷回运行时目录。Linux 安装会从源码构建 `pty.node`；CI 会在打包前进入匹配架构的 manylinux 2.28 容器重新构建该 addon，而 `--legacy` 部署会省略这一副作用目录，因此构建器会把它从根安装目录复制到暂存闭包。macOS 使用对应目标的预构建产物，并在可执行文件旁生成所需的 `-spawn-helper`。CI 将这些产物作为测试中间输入，只保留对应平台的 wheel 包。四个部署标志都有实测依据：未启用 `inject-workspace-packages` 时必须使用 `--legacy`；`hoisted` 为 pkg 提供稳定的单实例布局，再由显式物化步骤消除符号链接；关闭对等依赖自动安装可防止未声明的对等依赖扩大闭包；`link-workspace-packages` 选择直接工作区依赖。[`pnpm-workspace.yaml`](../../../../pnpm-workspace.yaml) 将传递的 `@deepseek-ai/cosmokit` 与 `@deepseek-ai/schemastery` semver 请求覆盖到固定的 vendor 源码，使 legacy deploy 不会从注册表解析这些未发布名称。
|
||||
[`scripts/build-exe-for-python-sdk.ts`](../../../../scripts/build-exe-for-python-sdk.ts)：运行时闭包校验 → `pnpm run build` →（清空后）`pnpm --filter dsh-jsonrpc-agent-pkg deploy --legacy --prod --config.node-linker=hoisted --config.auto-install-peers=false --config.link-workspace-packages=true` **直接写入** `python/sdk-runtime/src/deepseek_harness_runtime/runtime/node/` → 恢复被 legacy deploy 提升回源 manifest 的 `node_modules` 下的任何直接工作区包，同时省略其包内依赖树，并拒绝剩余的 manifest 缺口 → 将暂存依赖中的每个符号链接替换为目标文件内容，删除包管理器的 `.bin` 链接，并在仍有任何符号链接时失败 → 注入 pkg 配置（`bin` 指向闭包内的 `node_modules/@deepseek-ai/dsh-sdk-jsonrpc-demo/lib/packaged-bin.js`；`assets` 使用全量 glob，因为动态 `import()` 对 pkg 静态分析不可见，必须显式打入全部内容）→ 暂存目标平台的 `node-pty` addon → 每个构建目标调用一次 `pkg --sea` → 可执行文件 `dsh-jsonrpc-agent-pkg-<platform>-<arch>` 写入 `dist-exe/`，并拷回运行时目录。Linux 安装会从源码构建 `pty.node`；CI 会在打包前进入匹配架构的 manylinux 2.28 容器重新构建该 addon，而 `--legacy` 部署会省略这一副作用目录，因此构建器会把它从根安装目录复制到暂存闭包。macOS 使用对应目标的预构建产物，并在可执行文件旁生成所需的 `-spawn-helper`。CI 将这些产物作为测试中间输入，只保留对应平台的 wheel 包。四个部署标志都有实测依据：未启用 `inject-workspace-packages` 时必须使用 `--legacy`；`hoisted` 为 pkg 提供稳定的单实例布局，再由显式物化步骤消除符号链接；关闭对等依赖自动安装可防止未声明的对等依赖扩大闭包；`link-workspace-packages` 选择直接工作区依赖。[`pnpm-workspace.yaml`](../../../../pnpm-workspace.yaml) 将传递的 `@deepseek-ai/cosmokit` 与 `@deepseek-ai/schemastery` semver 请求覆盖到固定的 vendor 源码，使 legacy deploy 不会从注册表解析这些未发布名称。
|
||||
|
||||
CI 使用 [`.github/workflows/build-exe-for-python-sdk.yml`](../../../../.github/workflows/build-exe-for-python-sdk.yml)：[必需的 Python 运行时拉取请求验证](../testing/2026-08-12-required-python-runtime-pull-request-ci.md)调用它构建 linux-x64，手动派发 `workflow_dispatch` 或 PR（Pull Request）的 `build-exe` 标签可以显式选择构建目标，[公开发布工作流](../process/2026-08-11-python-publication-workflow.md)则调用它构建全部目标。linux-x64、linux-arm64（`ubuntu-24.04-arm`）和 macos-arm64 三个平台分别进行原生构建，并缓存 `~/.pkg-cache`；macOS 的 ad-hoc 签名由 pkg 处理。每个平台都使用 mock SSE（Server-Sent Events）模型，分别通过默认配置和自定义 `cordis.yml` 驱动 SDK，再通过 NDJSON JSON-RPC 直接驱动 exe，校验 JSONL 与最终响应；最后把发布形态的 wheel 包安装到干净的 venv 中，并在不传 `runtime_bin` 的情况下运行。Linux 还会检查可执行文件和原生 addon 各自的 GLIBC 依赖，并在 manylinux 2.28 容器中运行；macOS 则验证可执行文件的部署目标符合 wheel 包标签。完整构建三个目标时保留 4 个产物，每个产物只含一个发布文件：平台无关的 SDK wheel 包与 3 个原生运行时 wheel 包；手动选择部分目标时保留 SDK wheel 与所选运行时 wheel。裸 exe 与源码包只作为测试中间输入。[`.gitlab-ci.yml`](../../../../.gitlab-ci.yml) 只接受版本与根目录 `package.json` 匹配的 `python-v<repository-version>` 标签流水线，构建一个 SDK wheel 包和 3 个原生运行时 wheel 包，再由单个串行任务校验并将这 4 个文件发布到项目的 PyPI 注册表。Windows 不在目标范围内。
|
||||
|
||||
### Python SDK 分发：双载体，exe 用于生产，`node` 用于开发
|
||||
|
||||
Python SDK 位于 [`python/`](../../../../python/README.md)：`python/sdk` 是客户端，`python/sdk-runtime` 是运行时载体包。运行时包的数据目录包含检入的默认 `runtime/cordis.yml`、构建注入的平台 exe 与可选 helper，以及构建注入的 `runtime/node/` 闭包树。`resolve_bundled_launch_args()` 的自动解析**只查找 exe**；`node` 载体仅在显式设置 `DSH_RUNTIME_MODE=node` 时启用（运行 `runtime/node/node_modules/@deepseek-ai/dsh-jsonrpc-demo/lib/packaged-bin.js`，需要系统 Node ≥22.19），定位为本仓库成员的开发验证通道，不随 wheel 包分发。
|
||||
Python SDK 位于 [`python/`](../../../../python/README.md)：`python/sdk` 是客户端，`python/sdk-runtime` 是运行时载体包。运行时包的数据目录包含检入的默认 `runtime/cordis.yml`、构建注入的平台 exe 与可选 helper，以及构建注入的 `runtime/node/` 闭包树。`resolve_bundled_launch_args()` 的自动解析**只查找 exe**；`node` 载体仅在显式设置 `DSH_RUNTIME_MODE=node` 时启用（运行 `runtime/node/node_modules/@deepseek-ai/dsh-sdk-jsonrpc-demo/lib/packaged-bin.js`，需要系统 Node ≥22.19），定位为本仓库成员的开发验证通道，不随 wheel 包分发。
|
||||
|
||||
[`scripts/build-python-release.py`](../../../../scripts/build-python-release.py) 从仓库根目录的 `package.json` 读取权威的 `X.Y.Z` 或预发布版本，把预发布版本转换为 PEP 440 写法，并以该 wheel 包版本暂存两个包，让 `deepseek-harness-sdk` 精确依赖匹配版本的 `deepseek-harness-runtime-bin`。可选的 `python-v<repository-version>` 发布标签只是一项一致性断言，与仓库版本不同时会被拒绝；源码 `pyproject.toml` 中的开发占位版本从不决定发布版本。暂存过程还会把仓库许可证放入两个 wheel 包，并把第三方声明放入内置运行时 wheel 包。SDK 是 `py3-none-any` wheel 包；每个只提供 wheel 包的运行时包都包含一个 exe，macOS wheel 包还包含与其架构匹配的 helper。运行时 wheel 包使用 `py3-none-manylinux_2_28_x86_64`、`py3-none-manylinux_2_28_aarch64`，或针对 Node 24 可执行文件 macOS 13.5 部署目标而保守选择的 `py3-none-macosx_14_0_arm64` 标签；Hatch 钩子拒绝 sdist、通用标签、混合平台载荷、helper 缺失或多余，以及不支持的平台。
|
||||
|
||||
exe「必须显式配置」的硬语义不变；零配置体验由包装层恢复：调用方没有提供 `cordis`、没有显式指定运行时，且环境中没有 `DSH_CORDIS_CONFIG` 时，客户端将检入的默认 `cordis.yml`（`agent-core` + 预载的 `llm-deepseek` + JSONL 持久化 + `bash-local` + `dsh-jsonrpc` 对外服务条目，并通过 `!!js` 使用环境变量兜底）显式注入 `DSH_CORDIS_CONFIG`。
|
||||
exe「必须显式配置」的硬语义不变；零配置体验由包装层恢复：调用方没有提供 `cordis`、没有显式指定运行时，且环境中没有 `DSH_CORDIS_CONFIG` 时，客户端将检入的默认 `cordis.yml`（`agent-core` + 预载的 `llm-deepseek` + JSONL 持久化 + `bash-local` + `dsh-sdk-jsonrpc-server` 对外服务条目，并通过 `!!js` 使用环境变量兜底）显式注入 `DSH_CORDIS_CONFIG`。
|
||||
|
||||
### 命名血统
|
||||
|
||||
`@deepseek-ai/dsh-jsonrpc-demo`（包）→ `dsh-jsonrpc-agent`（`bin`）→ `dsh-jsonrpc-agent-pkg`（闭包 manifest；没有作用域前缀，刻意避开 `constraints` 对 `@deepseek-ai/dsh-*` 的包形状规则）→ `dsh-jsonrpc-agent-pkg-<platform>-<arch>`（exe 产物）。协议字段 `serverInfo.name` 保持为 `deepseek-harness-sdk-runtime`（协议稳定值）；Python 分发包名为 `deepseek-harness-sdk` / `deepseek-harness-runtime-bin`，导入模块名仍为 `deepseek_harness` / `deepseek_harness_runtime`。
|
||||
`@deepseek-ai/dsh-sdk-jsonrpc-demo`（包）→ `dsh-jsonrpc-agent`（`bin`）→ `dsh-jsonrpc-agent-pkg`（闭包 manifest；没有作用域前缀，刻意避开 `constraints` 对 `@deepseek-ai/dsh-*` 的包形状规则）→ `dsh-jsonrpc-agent-pkg-<platform>-<arch>`（exe 产物）。协议字段 `serverInfo.name` 保持为 `deepseek-harness-sdk-runtime`（协议稳定值）；Python 分发包名为 `deepseek-harness-sdk` / `deepseek-harness-runtime-bin`，导入模块名仍为 `deepseek_harness` / `deepseek_harness_runtime`。
|
||||
|
||||
## 工作线程插件
|
||||
|
||||
exe 内支持 `dsh-workflow-workerthread` 与 `dsh-code-runtime-worker`。两个后端构建后的宿主都通过 `fileURLToPath()` 转换相邻 `lib/worker.cjs` 的 URL，再将所得文件系统字符串传给 `Worker`；pkg 的 Worker 钩子可以用这种形式解析 VFS 内文件。该钩子会把 VFS 内的工作线程文件作为 CommonJS 编译，所以工作线程入口采用 CommonJS。工作流引擎在未构建的源码执行中仍保留 `data:` URL 引导程序，只有构建后的相邻入口使用文件系统字符串。自定义配置的可执行文件冒烟测试会加载两个后端，实际调用 `run_code` 与不启动 agent（智能体）的 `workflow`，并要求两个工作线程都从 pkg 的 VFS 内返回 `42`。
|
||||
exe 内支持 `dsh-workflow-worker-thread` 与 `dsh-code-runtime-worker-thread`。两个后端构建后的宿主都通过 `fileURLToPath()` 转换相邻 `lib/worker.cjs` 的 URL，再将所得文件系统字符串传给 `Worker`；pkg 的 Worker 钩子可以用这种形式解析 VFS 内文件。该钩子会把 VFS 内的工作线程文件作为 CommonJS 编译，所以工作线程入口采用 CommonJS。工作流引擎在未构建的源码执行中仍保留 `data:` URL 引导程序，只有构建后的相邻入口使用文件系统字符串。自定义配置的可执行文件冒烟测试会加载两个后端，实际调用 `run_code` 与不启动 agent（智能体）的 `workflow`，并要求两个工作线程都从 pkg 的 VFS 内返回 `42`。
|
||||
|
||||
## 测试
|
||||
|
||||
|
||||
@@ -2,5 +2,5 @@
|
||||
# side as of the last confirmed-consistent state. Both languages carry equal authority;
|
||||
# after editing either side, bring the other along and re-record with:
|
||||
# pnpm run verify-translation-pairing --write .agents/notes/implemented/architecture/2026-07-12-agent-scope-runtime-design.md
|
||||
2026-07-12-agent-scope-runtime-design.md: 5bee5f3f79be903848f8ecdf1ea84fe6fca60f6f
|
||||
2026-07-12-agent-scope-runtime-design.zh.md: b44969035a68ff62255a1179ba37cc30614d9521
|
||||
2026-07-12-agent-scope-runtime-design.md: 9d70b8048b1d2bb50290d34158d9deb329d5e15e
|
||||
2026-07-12-agent-scope-runtime-design.zh.md: a505f5337559294795c1080937c6079cee235baa
|
||||
|
||||
@@ -38,7 +38,7 @@ Five Cordis ideas are required to understand the implementation. A context selec
|
||||
|
||||
### A context is an ownership path through one service graph
|
||||
|
||||
All agents share one Cordis service graph. A derived context does not clone `ToolRegistry`, `SystemPrompt`, persistence, or model adapters; it changes how registrations made through that context are tagged and which effects own their cleanup.
|
||||
All agents share one Cordis service graph. A derived context does not clone `ToolRuntime`, `SystemPrompt`, persistence, or model adapters; it changes how registrations made through that context are tagged and which effects own their cleanup.
|
||||
|
||||
`agent.ctx` is such a derived context. Service calls still reach the shared instances, while a registration can inspect its calling context and store a contribution under the nearest scope key. Ordinary plugin contexts carry no scope key and therefore register globally.
|
||||
|
||||
@@ -226,7 +226,7 @@ After post-execute or outer pipeline normalization, the registry losslessly snap
|
||||
|
||||
SystemPrompt first resolves the global-plus-agent sections, variables, and tool providers into a deterministic registry contribution. The scope-filtered `system-prompt/assemble` waterfall may then reorder, replace, add, or remove any section, variable, or schema. Its returned assembly is authoritative; there is no later restoration pass and no finality metadata on ordinary prompt sections, tool definitions, or provider results.
|
||||
|
||||
This is a trusted same-process extension point, not an authority boundary. A listener that changes Code Mode's `run_code` schema or `tools:sdk` instructions, or a structured child's capture schema or instruction, owns preserving a coherent protocol in the assembly it returns. ToolRegistry still reserves `run_code` against ordinary tool registration and restriction because those are registry invariants, but assembly middleware remains free to transform the final model-visible surface.
|
||||
This is a trusted same-process extension point, not an authority boundary. A listener that changes Code Mode's `run_code` schema or `tools:sdk` instructions, or a structured child's capture schema or instruction, owns preserving a coherent protocol in the assembly it returns. ToolRuntime still reserves `run_code` against ordinary tool registration and restriction because those are registry invariants, but assembly middleware remains free to transform the final model-visible surface.
|
||||
|
||||
Scope solves the real isolation problem directly. Structured-output contributions register in the child's exact scope, while Code Mode derives its transport and SDK from the same resolved tool view. A second named-protection system would need another ownership and collision rule across arbitrary schema providers—including providers that intentionally contribute duplicate names—without creating a new trust boundary.
|
||||
|
||||
@@ -266,7 +266,7 @@ Subagent startup has one ownership transfer. The provider owns unpublished resou
|
||||
|
||||
### The service contract has one cancellation channel
|
||||
|
||||
`SubagentProvider.start()` and `SubagentService.start()` return `Promise<SubagentRun>`. The promise fulfills after the backend crosses its publication boundary, so callers and `subagent/start` observers never need a second `run.started` promise. Provider work that fails before publication rejects `start()`; prompt, turn, cancellation, and infrastructure outcomes after publication settle through `SubagentRun.result` without hiding the child id, as required by the [durable catalog decision](../feature/2026-07-22-durable-subagent-catalog-and-list-agents.md).
|
||||
`SubagentProvider.start()` and `SubagentRuntime.start()` return `Promise<SubagentRun>`. The promise fulfills after the backend crosses its publication boundary, so callers and `subagent/start` observers never need a second `run.started` promise. Provider work that fails before publication rejects `start()`; prompt, turn, cancellation, and infrastructure outcomes after publication settle through `SubagentRun.result` without hiding the child id, as required by the [durable catalog decision](../feature/2026-07-22-durable-subagent-catalog-and-list-agents.md).
|
||||
|
||||
`SubagentStartRequest.signal` is required. Aborting it requests cancellation during startup and across the published run's remaining readiness or turn work. `SubagentRun.dispose()` also requests cancellation and awaits quiescence. There is no separate public `run.cancel()` channel.
|
||||
|
||||
@@ -294,7 +294,7 @@ Worker and child-process bridges need more state than same-process registries be
|
||||
|
||||
### Workflow children are pending starts or published records
|
||||
|
||||
The workflow host keeps pending provider-start promises and published child records. A child moves from pending to published only when async `SubagentService.start()` fulfills; rejected starts clean their partial provider work and produce no child lifecycle pair.
|
||||
The workflow host keeps pending provider-start promises and published child records. A child moves from pending to published only when async `SubagentRuntime.start()` fulfills; rejected starts clean their partial provider work and produce no child lifecycle pair.
|
||||
|
||||
One host-owned AbortController supplies the required signal to pending and live children. Closing workflow admission aborts that signal, so there is no duplicate `ChildCancel` worker RPC or explicit host-side `run.cancel()` fanout. Quiescence waits for both pending starts and published child disposal.
|
||||
|
||||
@@ -376,7 +376,7 @@ The implementation is smaller and its proof follows the same shape as its owners
|
||||
- Create and resume expose no partially configured handle; final-entry losers and publication failures clean every prepared resource.
|
||||
- Disposal retains scoped listeners and persistence through driver drain and final session work, then revokes the scope.
|
||||
- Durable, queued, model, worker, process, and wire values are owned at their real boundary; typed same-process values follow readonly contracts.
|
||||
- ToolRegistry's presentation, lookup, and execution resolve the same live view before expert assembly transforms, and committed results have one immutable observation point.
|
||||
- ToolRuntime's presentation, lookup, and execution resolve the same live view before expert assembly transforms, and committed results have one immutable observation point.
|
||||
- Registry contributions are deterministic inputs, while the trusted assembly waterfall owns the final model-visible composition.
|
||||
- Subagent start returns only a published run, required signals cancel pending or live work, and disposal reaches the backend's quiescence contract.
|
||||
- Worker/process result precedence and cleanup remain correct under death, late messages, and bounded teardown.
|
||||
|
||||
@@ -38,7 +38,7 @@ Status: implemented
|
||||
|
||||
### 上下文是贯穿单个服务图的所有权路径
|
||||
|
||||
所有 agent 共享一个 Cordis 服务图。派生的上下文不会克隆 `ToolRegistry`、`SystemPrompt`、持久化或模型适配器；它改变的是：通过该上下文进行的注册如何被标记，以及哪些 effect 拥有其清理逻辑。
|
||||
所有 agent 共享一个 Cordis 服务图。派生的上下文不会克隆 `ToolRuntime`、`SystemPrompt`、持久化或模型适配器；它改变的是：通过该上下文进行的注册如何被标记，以及哪些 effect 拥有其清理逻辑。
|
||||
|
||||
`agent.ctx` 就是这样一个派生上下文。服务调用仍然到达共享实例，而注册操作可以检查其调用上下文并将贡献存储在最近的作用域键下。普通的插件上下文不携带作用域键，因此注册到全局。
|
||||
|
||||
@@ -226,7 +226,7 @@ Session 头部、种子和追加的事件是无损 JSON 数据。Session 构造
|
||||
|
||||
SystemPrompt 首先将全局加 agent 的段、变量和工具提供方解析为确定性的注册表贡献。作用域过滤的 `system-prompt/assemble` waterfall 随后可以重排、替换、添加或移除任何段、变量或 schema。其返回的组装结果即为权威；没有后续的恢复步骤，普通提示词段、工具定义或提供方结果上也没有终态元数据。
|
||||
|
||||
这是一个可信的同进程扩展点，而非权限边界。修改 Code Mode 的 `run_code` schema 或 `tools:sdk` 指令，或结构化子级的捕获 schema 或指令的监听器，有责任在其返回的组装中保持协议的一致性。ToolRegistry 仍然保留 `run_code` 不受普通工具注册和限制影响，因为那些是注册表不变式，但 assembly 中间件仍然可以自由变换最终的模型可见表面。
|
||||
这是一个可信的同进程扩展点，而非权限边界。修改 Code Mode 的 `run_code` schema 或 `tools:sdk` 指令，或结构化子级的捕获 schema 或指令的监听器，有责任在其返回的组装中保持协议的一致性。ToolRuntime 仍然保留 `run_code` 不受普通工具注册和限制影响，因为那些是注册表不变式，但 assembly 中间件仍然可以自由变换最终的模型可见表面。
|
||||
|
||||
Scope 直接解决了真正的隔离问题。结构化输出贡献注册在子级的精确作用域中，而 Code Mode 从同一个已解析的工具视图派生其传输和 SDK。第二套命名保护系统需要另一套所有权和碰撞规则来覆盖任意 schema 提供方（包括有意贡献重复名称的提供方），却不创建新的信任边界。
|
||||
|
||||
@@ -266,7 +266,7 @@ subagent 启动有一次所有权转移。提供方拥有未发布资源，直
|
||||
|
||||
### 服务约定有一个取消通道
|
||||
|
||||
`SubagentProvider.start()` 和 `SubagentService.start()` 返回 `Promise<SubagentRun>`。Promise 会在后端跨过发布边界后兑现，因此调用方和 `subagent/start` 观察者从不需要第二个 `run.started` promise。提供方工作如果在发布前失败，`start()` 就会被拒绝；发布后的提示词、轮次、取消与基础设施结果会通过 `SubagentRun.result` 结算，且不会隐藏 child id，这也是[持久化目录决策](../feature/2026-07-22-durable-subagent-catalog-and-list-agents.md)所要求的约定。
|
||||
`SubagentProvider.start()` 和 `SubagentRuntime.start()` 返回 `Promise<SubagentRun>`。Promise 会在后端跨过发布边界后兑现，因此调用方和 `subagent/start` 观察者从不需要第二个 `run.started` promise。提供方工作如果在发布前失败，`start()` 就会被拒绝；发布后的提示词、轮次、取消与基础设施结果会通过 `SubagentRun.result` 结算，且不会隐藏 child id，这也是[持久化目录决策](../feature/2026-07-22-durable-subagent-catalog-and-list-agents.md)所要求的约定。
|
||||
|
||||
`SubagentStartRequest.signal` 是必需的。中止它会在启动期间，以及已发布 run 的剩余就绪或轮次工作中请求取消。`SubagentRun.dispose()` 也请求取消并等待完全停稳。没有单独的公开 `run.cancel()` 通道。
|
||||
|
||||
@@ -294,7 +294,7 @@ Worker 和子进程桥接比同进程注册表需要更多状态，因为消息
|
||||
|
||||
### 工作流子级是待定 start 或已发布记录
|
||||
|
||||
工作流宿主保持待定的提供方 start promise 和已发布的子级记录。子级仅在异步 `SubagentService.start()` 兑现时才从待定变为已发布；被拒绝的 start 清理其部分提供方工作且不产生子级生命周期对。
|
||||
工作流宿主保持待定的提供方 start promise 和已发布的子级记录。子级仅在异步 `SubagentRuntime.start()` 兑现时才从待定变为已发布；被拒绝的 start 清理其部分提供方工作且不产生子级生命周期对。
|
||||
|
||||
一个宿主拥有的 AbortController 向待定和活跃子级提供必需的 signal。关闭工作流准入中止该 signal，因此没有重复的 `ChildCancel` worker RPC 或显式的宿主侧 `run.cancel()` 扇出。完全停稳需要等待待定 start 和已发布子级 dispose 两者。
|
||||
|
||||
@@ -376,7 +376,7 @@ Worker 消息、进程死亡和持久化输入确实跨越所有权和序列化
|
||||
- 创建和恢复不暴露部分配置的句柄；最终写入注册表时的失败者和发布失败清理每个已准备的资源。
|
||||
- dispose 在 driver 排空和最终会话工作期间保留作用域监听器和持久化，然后撤销作用域。
|
||||
- 持久化、队列、模型、worker、进程和协议格式的值在其真实边界处被拥有；类型化的同进程值遵循 readonly 约定。
|
||||
- ToolRegistry 的展示、查找和执行在专家 assembly 变换之前解析相同的活跃视图，已提交的结果有一个不可变的观察点。
|
||||
- ToolRuntime 的展示、查找和执行在专家 assembly 变换之前解析相同的活跃视图，已提交的结果有一个不可变的观察点。
|
||||
- 注册表贡献是确定性输入，而可信的 assembly waterfall 拥有最终的模型可见组合。
|
||||
- subagent start 仅返回已发布的 run，必需的 signal 取消待定或活跃的工作，dispose 到达后端的完全停稳约定。
|
||||
- Worker/进程结果优先级和清理在死亡、迟到消息和有界拆除下保持正确。
|
||||
|
||||
@@ -2,5 +2,5 @@
|
||||
# side as of the last confirmed-consistent state. Both languages carry equal authority;
|
||||
# after editing either side, bring the other along and re-record with:
|
||||
# pnpm run verify-translation-pairing --write .agents/notes/implemented/architecture/2026-07-12-scoped-layers-store.md
|
||||
2026-07-12-scoped-layers-store.md: 91df72cf18c3f28d49534793bee85094b299c9a5
|
||||
2026-07-12-scoped-layers-store.zh.md: 6ff44b76ee3df731be02c98e9ab8bd484c428055
|
||||
2026-07-12-scoped-layers-store.md: 221711748007d5f909aec3fdfcf3919884d11cdb
|
||||
2026-07-12-scoped-layers-store.zh.md: 7f3ad11e23f02eeafe955efe08525bda8a9e9996
|
||||
|
||||
@@ -6,7 +6,7 @@ English | [中文](2026-07-12-scoped-layers-store.zh.md)
|
||||
|
||||
## Problem
|
||||
|
||||
Agent scoping ([decision](2026-07-08-agent-scope-contexts.md), [runtime design](2026-07-12-agent-scope-runtime-design.md)) gives scope-aware registries the same recurring shape: one global registration layer plus one exact agent layer. Seven registration facades use that shape: `tools.register`, `tools.restrict`, and `tools.guard` in `dsh-tools`; `SystemPrompt.section`, `SystemPrompt.tools`, and `SystemPrompt.variable` in `dsh-system-prompt`; and `CommandService.register` in `dsh-commands`.
|
||||
Agent scoping ([decision](2026-07-08-agent-scope-contexts.md), [runtime design](2026-07-12-agent-scope-runtime-design.md)) gives scope-aware registries the same recurring shape: one global registration layer plus one exact agent layer. Seven registration facades use that shape: `tools.register`, `tools.restrict`, and `tools.guard` in `dsh-tools`; `SystemPrompt.section`, `SystemPrompt.tools`, and `SystemPrompt.variable` in `dsh-system-prompt`; and `CommandRuntime.register` in `dsh-commands`.
|
||||
|
||||
Without a shared primitive, each facade repeats the lifecycle choreography around its domain state: derive visibility from the calling context, create a scoped container on demand, attach ownership to the same Cordis fiber, install undo before notifying observers, return Cordis's exact disposer, and reclaim empty scoped state. Separate maps and collection types also leave a service without one object representing a scope's complete contribution.
|
||||
|
||||
@@ -78,15 +78,15 @@ export class AnonymousEntries<V> {
|
||||
- `AnonymousEntries.append()` assigns a unique internal key per registration, so equal callbacks or values remain independent. Its iterator is insertion-ordered and uses the same live-generation boundary.
|
||||
- `effect()` derives the key with `scopeOf(ctx)` and attaches the action to that same `ctx.effect()`. It accepts one synchronous action returning one synchronous undo; actions must either return their undo or throw before retaining a contribution. The helper does not normalize the wider Cordis `Effect` union.
|
||||
- `effect()` collects the action's undo before calling `onChange` and returns the exact `ctx.effect()` disposer. Disposal runs the action undo before notification, is idempotent through Cordis, and removes a scoped layer only after its complete `ScopeLayer.isEmpty()` becomes true.
|
||||
- `options.notify` defaults to `true`. The callback's own policy stays authoritative: tool and prompt change callbacks may throw and trigger registration rollback; `CommandService.notifyChange()` contains observer failures; tool guards pass `notify: false`.
|
||||
- `options.notify` defaults to `true`. The callback's own policy stays authoritative: tool and prompt change callbacks may throw and trigger registration rollback; `CommandRuntime.notifyChange()` contains observer failures; tool guards pass `notify: false`.
|
||||
|
||||
## Registry migrations
|
||||
|
||||
`dsh-tools` defines one `ToolLayer` containing named tools plus anonymous compiled restrictions and guard registrations. `ToolRegistry` retains its private domain resolver for visible definitions, pre-restriction known names, restrictable global names, scoped shadowing, restrictions, and reserved `run_code` insertion. Guard evaluation live-iterates global then scoped registrations: additions to a nonempty generation can run in the current dispatch, while a self-replacement after draining the guard table begins with the next dispatch.
|
||||
`dsh-tools` defines one `ToolLayer` containing named tools plus anonymous compiled restrictions and guard registrations. `ToolRuntime` retains its private domain resolver for visible definitions, pre-restriction known names, restrictable global names, scoped shadowing, restrictions, and reserved `run_code` insertion. Guard evaluation live-iterates global then scoped registrations: additions to a nonempty generation can run in the current dispatch, while a self-replacement after draining the guard table begins with the next dispatch.
|
||||
|
||||
`dsh-system-prompt` defines one `PromptLayer` containing named sections and variables plus anonymous tool providers. Assembly merges sections before evaluating them, so a shadowed provider is never called. Tool-provider membership is materialized once per assembly. Variable providers live-iterate global then scoped tables: additions to a nonempty generation can run in the current assembly, while a self-replacement after draining the variable table begins with the next assembly.
|
||||
|
||||
`dsh-commands` defines a one-table layer containing `NamedEntries<RegisteredCommand>`. Effective views use `merge()`, while `CommandService` retains definition normalization and freezing, exact duplicate diagnostics, sorted immutable descriptors, direct execution, HMR cleanup, and independently contained `commands/change` observers.
|
||||
`dsh-commands` defines a one-table layer containing `NamedEntries<RegisteredCommand>`. Effective views use `merge()`, while `CommandRuntime` retains definition normalization and freezing, exact duplicate diagnostics, sorted immutable descriptors, direct execution, HMR cleanup, and independently contained `commands/change` observers.
|
||||
|
||||
All seven facades keep validation and diagnostics in their owning registry and continue to return the exact Cordis disposer. The migration changes neither public registry behavior nor model-, human-, wire-, persistence-, or configuration-visible output.
|
||||
|
||||
@@ -102,7 +102,7 @@ All seven facades keep validation and diagnostics in their owning registry and c
|
||||
|
||||
**Accept the complete Cordis `Effect` union.** None of the seven registrations has asynchronous setup, multiple undos, or an independent settlement boundary. General normalization would duplicate Cordis lifecycle machinery without a current consumer.
|
||||
|
||||
**Expose `ScopedLayers.values()`, `ScopedLayers.keys()`, or a global-admission predicate.** Those operations encode consumer-specific live/materialized and filtering policies. Direct table iteration preserves explicit live semantics, `merge()` covers the shared named shadowing operation, and `ToolRegistry` keeps its richer private resolver.
|
||||
**Expose `ScopedLayers.values()`, `ScopedLayers.keys()`, or a global-admission predicate.** Those operations encode consumer-specific live/materialized and filtering policies. Direct table iteration preserves explicit live semantics, `merge()` covers the shared named shadowing operation, and `ToolRuntime` keeps its richer private resolver.
|
||||
|
||||
**Put `values()` on `ScopeLayer` or export `EntryValues`.** A layer aggregates heterogeneous tables and has no coherent value type or iteration policy. `EntryValues` is useful only to share implementation details between the two table classes; making it public would enlarge the interface without giving callers a meaningful layer-wide read.
|
||||
|
||||
|
||||
@@ -6,7 +6,7 @@ Status: implemented
|
||||
|
||||
## 问题
|
||||
|
||||
agent（智能体）作用域机制（[决策](2026-07-08-agent-scope-contexts.md)、[运行时设计](2026-07-12-agent-scope-runtime-design.md)）让支持作用域的注册表反复呈现同一种形态：一个全局注册层，加上一个与具体 agent 精确对应的层。七个注册门面都采用这一形态：`tools.register`、`tools.restrict` 和 `tools.guard`（位于 `dsh-tools`）；`SystemPrompt.section`、`SystemPrompt.tools` 和 `SystemPrompt.variable`（位于 `dsh-system-prompt`）；以及 `CommandService.register`（位于 `dsh-commands`）。
|
||||
agent（智能体）作用域机制（[决策](2026-07-08-agent-scope-contexts.md)、[运行时设计](2026-07-12-agent-scope-runtime-design.md)）让支持作用域的注册表反复呈现同一种形态：一个全局注册层，加上一个与具体 agent 精确对应的层。七个注册门面都采用这一形态：`tools.register`、`tools.restrict` 和 `tools.guard`（位于 `dsh-tools`）；`SystemPrompt.section`、`SystemPrompt.tools` 和 `SystemPrompt.variable`（位于 `dsh-system-prompt`）；以及 `CommandRuntime.register`（位于 `dsh-commands`）。
|
||||
|
||||
如果没有共享原语，每个门面都要围绕自己的领域状态重复相同的生命周期编排：从调用方上下文导出可见性，按需创建专属容器，把属主绑定到同一个 Cordis fiber，先装入 undo 再通知观察者，原样返回 Cordis 的 disposer，并回收空的专属状态。各自分离的映射与集合类型也会让服务缺少一个表示某个 scope 完整贡献的对象。
|
||||
|
||||
@@ -78,15 +78,15 @@ export class AnonymousEntries<V> {
|
||||
- `AnonymousEntries.append()` 为每次登记分配唯一内部键，因此值相等的回调或其他值仍彼此独立。其迭代器保留插入顺序，并采用同样的 generation 活遍历边界。
|
||||
- `effect()` 通过 `scopeOf(ctx)` 导出键，并把 action 挂到同一个 `ctx.effect()` 上。它只接受一个同步 action，且该 action 只返回一个同步 undo；action 要么返回其 undo，要么必须在保留任何贡献之前抛错。helper 不会规范化更宽泛的 Cordis `Effect` union。
|
||||
- `effect()` 在调用 `onChange` 前收集 action 的 undo，并原样返回 `ctx.effect()` 的 disposer。销毁时先运行 action undo 再通知；Cordis 保证其幂等性；只有在整个层的 `ScopeLayer.isEmpty()` 返回 true 后，helper 才会删除专属层。
|
||||
- `options.notify` 默认为 `true`。回调自身的策略仍具最终效力：工具与提示词的 change 回调可以抛错并触发登记回滚；`CommandService.notifyChange()` 会隔离观察者失败；工具 guard 传入 `notify: false`。
|
||||
- `options.notify` 默认为 `true`。回调自身的策略仍具最终效力：工具与提示词的 change 回调可以抛错并触发登记回滚；`CommandRuntime.notifyChange()` 会隔离观察者失败；工具 guard 传入 `notify: false`。
|
||||
|
||||
## 注册表迁移
|
||||
|
||||
`dsh-tools` 定义一个 `ToolLayer`，其中包含命名工具以及匿名的已编译 restriction 和 guard 登记。`ToolRegistry` 保留其私有领域解析器，由它处理可见定义、限制前的已知名称、可限制的全局名称、专属遮蔽、restriction，以及保留的 `run_code` 插入。guard 求值会先活遍历全局登记，再活遍历专属登记：向非空 generation 新增的登记可以在当前分发中运行，而 guard 表清空后的自我替换则从下一次分发开始运行。
|
||||
`dsh-tools` 定义一个 `ToolLayer`，其中包含命名工具以及匿名的已编译 restriction 和 guard 登记。`ToolRuntime` 保留其私有领域解析器，由它处理可见定义、限制前的已知名称、可限制的全局名称、专属遮蔽、restriction，以及保留的 `run_code` 插入。guard 求值会先活遍历全局登记，再活遍历专属登记：向非空 generation 新增的登记可以在当前分发中运行，而 guard 表清空后的自我替换则从下一次分发开始运行。
|
||||
|
||||
`dsh-system-prompt` 定义一个 `PromptLayer`，其中包含命名的段落与变量，以及匿名工具提供方。组装流程在求值前合并段落，因此被遮蔽的提供方不会被调用。每次组装只物化一次工具提供方成员集合。变量提供方会先活遍历全局表，再活遍历专属表：向非空 generation 新增的提供方可以在当前组装中运行，而变量表清空后的自我替换则从下一次组装开始运行。
|
||||
|
||||
`dsh-commands` 定义一个单表层，其中包含 `NamedEntries<RegisteredCommand>`。生效视图使用 `merge()`；`CommandService` 则保留对定义的规范化与冻结处理、精确重名诊断、经过排序的不可变描述符、直接执行、HMR（热模块替换）清理，以及对各个 `commands/change` 观察者分别隔离失败的行为。
|
||||
`dsh-commands` 定义一个单表层，其中包含 `NamedEntries<RegisteredCommand>`。生效视图使用 `merge()`；`CommandRuntime` 则保留对定义的规范化与冻结处理、精确重名诊断、经过排序的不可变描述符、直接执行、HMR（热模块替换）清理，以及对各个 `commands/change` 观察者分别隔离失败的行为。
|
||||
|
||||
七个门面都把校验与诊断留在所属注册表中，并继续返回 Cordis 的原始 disposer。迁移既不改变公开注册表行为，也不改变模型可见或人类可见的输出，以及协议、持久化或配置层面的可见输出。
|
||||
|
||||
@@ -102,7 +102,7 @@ export class AnonymousEntries<V> {
|
||||
|
||||
**接受完整的 Cordis `Effect` union。** 七个登记口都不涉及异步 setup、多份 undo 或独立结算边界。若没有现有消费方需要，通用规范化只会重复实现 Cordis 的生命周期机制。
|
||||
|
||||
**暴露 `ScopedLayers.values()`、`ScopedLayers.keys()` 或全局放行谓词。** 这些操作会编码消费方特有的活遍历或物化策略，以及过滤策略。直接遍历条目表可保留显式的活语义，`merge()` 覆盖共享的命名遮蔽操作，而 `ToolRegistry` 继续保有功能更丰富的私有解析器。
|
||||
**暴露 `ScopedLayers.values()`、`ScopedLayers.keys()` 或全局放行谓词。** 这些操作会编码消费方特有的活遍历或物化策略，以及过滤策略。直接遍历条目表可保留显式的活语义，`merge()` 覆盖共享的命名遮蔽操作，而 `ToolRuntime` 继续保有功能更丰富的私有解析器。
|
||||
|
||||
**把 `values()` 放在 `ScopeLayer` 上，或导出 `EntryValues`。** 一个层会聚合异构表，因而没有一致的值类型或迭代策略。`EntryValues` 只适合在两个表类之间共享实现细节；将其公开只会扩大接口，却不能为调用方提供有意义的整层读取方式。
|
||||
|
||||
|
||||
@@ -2,5 +2,5 @@
|
||||
# side as of the last confirmed-consistent state. Both languages carry equal authority;
|
||||
# after editing either side, bring the other along and re-record with:
|
||||
# pnpm run verify-translation-pairing --write .agents/notes/implemented/architecture/2026-07-14-provider-routed-llm-adapters.md
|
||||
2026-07-14-provider-routed-llm-adapters.md: df152a4436226ac5cb1941ce060a35bd4a0d496d
|
||||
2026-07-14-provider-routed-llm-adapters.zh.md: 04fb1082942e777c9ea1a95cea928ccd788d24df
|
||||
2026-07-14-provider-routed-llm-adapters.md: e1eaf52f21481a7c65e85effb7607b16f9b0ffdd
|
||||
2026-07-14-provider-routed-llm-adapters.zh.md: 4e620408dabffc8368293635613afb5778b6e822
|
||||
|
||||
@@ -6,7 +6,7 @@ English | [中文](2026-07-14-provider-routed-llm-adapters.zh.md)
|
||||
|
||||
## Problem
|
||||
|
||||
`dsh-llm` registered adapters by exact model name. A plugin supplied a model list at Cordis startup, `LlmService` stored one adapter per listed string, and `GenerateOptions.model` selected the adapter and the provider model at once. This worked while both shipping adapters targeted the same two DeepSeek models, but it conflated two independent decisions: which upstream provider owns a request, and which model that provider should run.
|
||||
`dsh-llm` registered adapters by exact model name. A plugin supplied a model list at Cordis startup, `LlmRuntime` stored one adapter per listed string, and `GenerateOptions.model` selected the adapter and the provider model at once. This worked while both shipping adapters targeted the same two DeepSeek models, but it conflated two independent decisions: which upstream provider owns a request, and which model that provider should run.
|
||||
|
||||
The conflation prevents a provider gateway from serving an open-ended model catalog. OpenRouter, for example, is one provider with many model ids, while a private OpenAI-compatible endpoint may add models without changing the Harness plugin tree. Every newly selected model currently needs to have been registered during plugin startup. The same model id can also exist at multiple providers, so model-only registration cannot state which provider the caller intended.
|
||||
|
||||
@@ -20,7 +20,7 @@ The adapter configuration also assumes one DeepSeek API key and endpoint. A gene
|
||||
|
||||
`GenerateOptions` and `LlmCallConfig` carry `provider: string` beside `model: string`; `AgentOptions` carries the corresponding optional creation field. A loop request is valid only after both values are non-empty, and both values are part of the logged request header. `agent/request` may return a replacement pair on any step, so a session can switch providers and models without changing the Cordis plugin lifecycle.
|
||||
|
||||
`LlmService` registers and resolves adapters by provider. `registerAdapter(providers, adapter)` checks the entire provider list before mutating the registry, rejects a duplicate with `DUPLICATE_ADAPTER`, and disposes the whole registration as one effect. Model ids are not registration keys; the selected adapter still validates or forwards them. The later [LLM catalog and ACP selection Agent Note](2026-07-15-llm-model-catalog-and-acp-selection.md) added advisory `listProviders()` / `listModels()` discovery without turning model membership into request validation.
|
||||
`LlmRuntime` registers and resolves adapters by provider. `registerAdapter(providers, adapter)` checks the entire provider list before mutating the registry, rejects a duplicate with `DUPLICATE_ADAPTER`, and disposes the whole registration as one effect. Model ids are not registration keys; the selected adapter still validates or forwards them. The later [LLM catalog and ACP selection Agent Note](2026-07-15-llm-model-catalog-and-acp-selection.md) added advisory `listProviders()` / `listModels()` discovery without turning model membership into request validation.
|
||||
|
||||
A provider has exactly one adapter owner in a Cordis context. `dsh-llm-deepseek` registers `deepseek`; `dsh-llm-pi-ai` may also register `deepseek`, but loading both owners is a configuration error rather than an ordering rule or fallback. A deployment that wants the hand-rolled DeepSeek implementation excludes `deepseek` from the pi-ai profiles. A deployment that wants pi-ai's DeepSeek implementation does not mount `dsh-llm-deepseek`.
|
||||
|
||||
@@ -42,7 +42,7 @@ Assistant messages carry the request's `provider` and `model`, plus an optional
|
||||
|
||||
A terminal successful `finish` chunk may carry replay state, and `BlockAssembler` retains it alongside usage and finish reason. The loop attaches that state to the assembled assistant message's model source without exposing a response-rewrite hook. Error and aborted responses do not produce a normal assistant message and therefore do not enter future model history.
|
||||
|
||||
The pi-ai replay state is a versioned, minimal projection of its successful `AssistantMessage`: source API/provider/model, response id/model, stop reason, and index-aligned text, thinking, and tool-call signatures. It does not duplicate text or tool arguments already carried by Harness content blocks, and it omits diagnostics, timestamps, usage, and errors. On a later request, `LlmService` gives replay state to the target adapter only when the historical provider and target provider are currently owned by the same adapter instance. That adapter combines the logged Harness content with replay state when it can restore the historical response, and owns any required cross-model or cross-provider conversion. An adapter receiving replay state with an unknown version or mismatched block shape fails explicitly; a different adapter receives only provider-neutral content plus provider/model fields.
|
||||
The pi-ai replay state is a versioned, minimal projection of its successful `AssistantMessage`: source API/provider/model, response id/model, stop reason, and index-aligned text, thinking, and tool-call signatures. It does not duplicate text or tool arguments already carried by Harness content blocks, and it omits diagnostics, timestamps, usage, and errors. On a later request, `LlmRuntime` gives replay state to the target adapter only when the historical provider and target provider are currently owned by the same adapter instance. That adapter combines the logged Harness content with replay state when it can restore the historical response, and owns any required cross-model or cross-provider conversion. An adapter receiving replay state with an unknown version or mismatched block shape fails explicitly; a different adapter receives only provider-neutral content plus provider/model fields.
|
||||
|
||||
This state is model-visible replay input and therefore follows the existing [reconstructable-request rule](2026-07-05-reconstructable-requests.md): it is present in both the terminal `finish` chunk and the assembled `assistant/message` model source that drives derivation. Resume and fork preserve it verbatim. Compaction that shadows the assistant message also removes its replay state from the active surface; the summary is ordinary provider-neutral content.
|
||||
|
||||
@@ -50,7 +50,7 @@ This state is model-visible replay input and therefore follows the existing [rec
|
||||
|
||||
Every model-selection path carries provider and model together: declarative agents, ACP and stdio app config, the JSON-RPC initialize request, subagent overrides and inheritance, workflow child overrides, and direct compaction summarization. Subagents inherit both fields from their parent before applying request overrides. The system-prompt variable set gains `provider` beside `model`.
|
||||
|
||||
Compaction configuration gains `summarizationProvider` beside `summarizationModel`. Both are empty to inherit, or both are non-empty to select an explicit target; a half-configured pair fails load. Inheritance uses the last logged request target when one exists and falls back to the agent's creation options. `compact/summary` records both fields with the existing model-call envelope.
|
||||
Compaction configuration gains `summarizationProvider` beside `summarizationModel`. Both are empty to inherit, or both are non-empty to select an explicit target; a half-configured pair fails load. Inheritance uses the last logged request target when one exists and falls back to the agent's creation options. `compaction/summary` records both fields with the existing model-call envelope.
|
||||
|
||||
The JSON-RPC runtime receives provider and model explicitly. Its convenience fallback mounts `dsh-llm-deepseek` only for provider `deepseek` when that provider has no registered owner; other missing providers fail without guessing an adapter.
|
||||
|
||||
|
||||
@@ -6,7 +6,7 @@ Status: implemented
|
||||
|
||||
## 问题
|
||||
|
||||
`dsh-llm` 按精确模型名称注册适配器。插件在 Cordis 启动时提供模型列表，`LlmService` 为列表中的每个字符串保存一个适配器，`GenerateOptions.model` 同时选择适配器与提供方模型。两个随附的适配器都只面向相同的两个 DeepSeek 模型时，这种方式可以工作，但它混淆了两个独立决策：由哪个上游提供方承接请求，以及该提供方应运行哪个模型。
|
||||
`dsh-llm` 按精确模型名称注册适配器。插件在 Cordis 启动时提供模型列表，`LlmRuntime` 为列表中的每个字符串保存一个适配器，`GenerateOptions.model` 同时选择适配器与提供方模型。两个随附的适配器都只面向相同的两个 DeepSeek 模型时，这种方式可以工作，但它混淆了两个独立决策：由哪个上游提供方承接请求，以及该提供方应运行哪个模型。
|
||||
|
||||
这种混淆使提供方网关无法提供开放的模型目录。例如，OpenRouter 是一个包含大量模型 ID 的提供方，私有 OpenAI 兼容端点也可能在不修改 Harness 插件树的情况下增加模型。目前，每个新选择的模型都必须在插件启动期间完成注册。同一个模型 ID 还可能存在于多个提供方中，因此仅按模型注册无法表达调用方预期使用的提供方。
|
||||
|
||||
@@ -20,7 +20,7 @@ Status: implemented
|
||||
|
||||
`GenerateOptions` 与 `LlmCallConfig` 在 `model: string` 之外携带 `provider: string`，`AgentOptions` 则携带对应的可选创建字段。只有两个值都非空时，agent loop（智能体循环）请求才有效；两个值也都会写入请求头日志。`agent/request` 可以在任意步骤返回替换后的字段组合，因此会话可以切换提供方与模型，无需改变 Cordis 插件生命周期。
|
||||
|
||||
`LlmService` 按提供方注册和解析适配器。`registerAdapter(providers, adapter)` 在修改注册表前检查整个提供方列表，遇到重复项时返回 `DUPLICATE_ADAPTER`，并以一个 effect 为单位整体 dispose（资源释放）。模型 ID 不作为注册键；仍由选中的适配器负责验证或转发。后续的 [LLM 目录与 ACP 模型选择 Agent Note](2026-07-15-llm-model-catalog-and-acp-selection.md) 增加了建议性的 `listProviders()` / `listModels()` 发现接口，但不会把目录成员关系变成请求校验规则。
|
||||
`LlmRuntime` 按提供方注册和解析适配器。`registerAdapter(providers, adapter)` 在修改注册表前检查整个提供方列表，遇到重复项时返回 `DUPLICATE_ADAPTER`，并以一个 effect 为单位整体 dispose（资源释放）。模型 ID 不作为注册键；仍由选中的适配器负责验证或转发。后续的 [LLM 目录与 ACP 模型选择 Agent Note](2026-07-15-llm-model-catalog-and-acp-selection.md) 增加了建议性的 `listProviders()` / `listModels()` 发现接口，但不会把目录成员关系变成请求校验规则。
|
||||
|
||||
在一个 Cordis 上下文中，一个提供方只能有一个适配器所有者。`dsh-llm-deepseek` 注册 `deepseek`；`dsh-llm-pi-ai` 也可以注册 `deepseek`，但同时加载两个所有者属于配置错误，不采用顺序规则或回退行为。若部署选择手写的 DeepSeek 实现，需从 pi-ai 配置中排除 `deepseek`；若部署选择 pi-ai 的 DeepSeek 实现，则不挂载 `dsh-llm-deepseek`。
|
||||
|
||||
@@ -42,7 +42,7 @@ pi-ai 的通用流选项不支持停止序列。若 Harness `stop` 选项已定
|
||||
|
||||
成功的终止 `finish` 分片可以携带回放状态，`BlockAssembler` 会将其与 token 用量和结束原因一起保留。agent loop 会把该状态附加到已组装助手消息的模型来源中，但不公开响应改写钩子。错误或中止响应不会生成正常助手消息，因此不会进入后续模型历史。
|
||||
|
||||
pi-ai 回放状态是其成功 `AssistantMessage` 的带版本最小投影，包含源 API/提供方/模型、响应 ID/模型、停止原因，以及按索引对齐的文本签名、thinking 签名和工具调用签名。它不会重复 Harness 内容块中已有的文本或工具参数，也不包含诊断信息、时间戳、用量或错误。后续请求中，只有历史提供方和目标提供方当前归同一个适配器实例所有时，`LlmService` 才会把回放状态交给目标适配器。适配器在能够恢复历史响应时，将 Harness 记录的内容与回放状态组合，并负责所需的跨模型或跨提供方转换。适配器收到未知版本或块形状不匹配的回放状态时会显式失败；其他适配器只能收到提供方无关的内容以及提供方/模型字段。
|
||||
pi-ai 回放状态是其成功 `AssistantMessage` 的带版本最小投影，包含源 API/提供方/模型、响应 ID/模型、停止原因，以及按索引对齐的文本签名、thinking 签名和工具调用签名。它不会重复 Harness 内容块中已有的文本或工具参数，也不包含诊断信息、时间戳、用量或错误。后续请求中，只有历史提供方和目标提供方当前归同一个适配器实例所有时，`LlmRuntime` 才会把回放状态交给目标适配器。适配器在能够恢复历史响应时，将 Harness 记录的内容与回放状态组合，并负责所需的跨模型或跨提供方转换。适配器收到未知版本或块形状不匹配的回放状态时会显式失败；其他适配器只能收到提供方无关的内容以及提供方/模型字段。
|
||||
|
||||
该状态属于模型可见的回放输入，因此遵循现有的[请求可重建规则](2026-07-05-reconstructable-requests.md)：它同时存在于终止 `finish` 分片和驱动派生的已组装 `assistant/message` 模型来源中。恢复和 fork 会原样保留该状态。压缩（compaction）遮蔽助手消息时，也会从活动 surface 中移除其回放状态；摘要属于普通的提供方无关内容。
|
||||
|
||||
@@ -50,7 +50,7 @@ pi-ai 回放状态是其成功 `AssistantMessage` 的带版本最小投影，包
|
||||
|
||||
每条模型选择路径都同时携带 provider 与 model：声明式 agent、ACP（Agent Client Protocol）和 stdio 应用配置、JSON-RPC initialize 请求、subagent 覆盖与继承、工作流子 agent 覆盖，以及直接压缩摘要。subagent 先从父 agent 继承两个字段，再应用请求覆盖。系统提示词变量集合在 `model` 之外增加 `provider`。
|
||||
|
||||
压缩配置在 `summarizationModel` 之外增加 `summarizationProvider`。两个值均为空时继承，均非空时选择显式目标；只配置其中一个会导致加载失败。继承优先使用最近一次记录的请求目标，没有时回退到 agent 创建选项。`compact/summary` 使用现有模型调用 envelope 记录两个字段。
|
||||
压缩配置在 `summarizationModel` 之外增加 `summarizationProvider`。两个值均为空时继承，均非空时选择显式目标；只配置其中一个会导致加载失败。继承优先使用最近一次记录的请求目标，没有时回退到 agent 创建选项。`compaction/summary` 使用现有模型调用 envelope 记录两个字段。
|
||||
|
||||
JSON-RPC 运行时显式接收提供方与模型。仅当 `deepseek` 提供方没有注册所有者时，其便利回退才会挂载 `dsh-llm-deepseek`；其他缺失的提供方会直接失败，不会猜测适配器。
|
||||
|
||||
|
||||
@@ -2,5 +2,5 @@
|
||||
# side as of the last confirmed-consistent state. Both languages carry equal authority;
|
||||
# after editing either side, bring the other along and re-record with:
|
||||
# pnpm run verify-translation-pairing --write .agents/notes/implemented/architecture/2026-07-15-agent-initiator-scope.md
|
||||
2026-07-15-agent-initiator-scope.md: 2f388ae1de3dd6583686e129a03f0cd76701e18d
|
||||
2026-07-15-agent-initiator-scope.md: 63540c0ec6b29a10613e01f1ed9ced24e8f2d277
|
||||
2026-07-15-agent-initiator-scope.zh.md: a0e3638081d875adc2412191829ab84c8dcd697c
|
||||
|
||||
@@ -20,7 +20,7 @@ The mandatory `ctx.agents` service uses Node `AsyncLocalStorage` to carry the in
|
||||
|
||||
Concurrent drivers receive independent stores. A child driver's continuations carry the child, while the caller resumes in its prior store as soon as `withInitiator()` returns; active-run tracking keeps the returned Promise in the teardown drain until it settles. Creation, persistence load, and unpublished `setup(agentCtx)` remain outside the child's driver boundary: creation initiated by a parent runs under the parent identity, while `agentCtx.agent` explicitly identifies the child.
|
||||
|
||||
Ambient identity does not replace explicit contracts. `ToolExecution.agent`, `AssembleContext.agent`, `GenerateOptions.sessionId`, task ownership, parent/child requests, `ctx.agent`, `agentCtx.agent`, approval and hook subjects, `cwd` selection, cancellation, worker/process messages, persistence records, and wire identity remain explicit. A remote boundary materializes the identity it needs into its typed request because ALS is process-local.
|
||||
Ambient identity does not replace explicit contracts. `ToolExecution.agent`, `AssembleContext.agent`, `GenerateOptions.sessionId`, job ownership, parent/child requests, `ctx.agent`, `agentCtx.agent`, approval and hook subjects, `cwd` selection, cancellation, worker/process messages, persistence records, and wire identity remain explicit. A remote boundary materializes the identity it needs into its typed request because ALS is process-local.
|
||||
|
||||
`AgentRegistry` owns an ordered initiator lifecycle. Teardown first rejects new boundaries; removing `ctx.agents` then drains injected dependents such as AgentLoop, and the registry waits for active returned-Promise boundaries before calling `AsyncLocalStorage.disable()`. If a boundary's inherited async chain starts an owning Cordis fiber's unload, the private run-token lineage releases that nested boundary chain from the drain, which prevents teardown from waiting on itself while unrelated boundaries still drain. `currentInitiator()` and `requireInitiator()` remain usable through a retained in-flight service reference while the ordinary drain runs; after disposal, initiator methods throw `agent initiator scope is disposed`. Root Context disposal may start sibling fiber teardown concurrently, so active-boundary counting remains necessary in addition to Cordis dependency ordering.
|
||||
|
||||
|
||||
@@ -2,5 +2,5 @@
|
||||
# side as of the last confirmed-consistent state. Both languages carry equal authority;
|
||||
# after editing either side, bring the other along and re-record with:
|
||||
# pnpm run verify-translation-pairing --write .agents/notes/implemented/architecture/2026-07-15-llm-model-catalog-and-acp-selection.md
|
||||
2026-07-15-llm-model-catalog-and-acp-selection.md: bfd17c73b01319c10d5dc03333b3c726db5d6f33
|
||||
2026-07-15-llm-model-catalog-and-acp-selection.zh.md: 2ee87348c4719d304e7b363b4c7bf7522632962b
|
||||
2026-07-15-llm-model-catalog-and-acp-selection.md: 3dddbe7e9fae74e4ce1ec1c8e93a40c3352b5a54
|
||||
2026-07-15-llm-model-catalog-and-acp-selection.zh.md: 145fd0bc379ff8132d09ec628b6381a18f755cfa
|
||||
|
||||
@@ -8,7 +8,7 @@ English | [中文](2026-07-15-llm-model-catalog-and-acp-selection.zh.md)
|
||||
|
||||
## Problem
|
||||
|
||||
Provider-routed adapters let every request choose `provider + model`, but `LlmService` exposed only routing and streaming. A UI could not discover which providers were registered or which models an adapter was prepared to recommend. ACP clients therefore received no `model` session config option, so Zed, JetBrains, and VS Code integrations had no model list even though the LLM service already supported runtime switching.
|
||||
Provider-routed adapters let every request choose `provider + model`, but `LlmRuntime` exposed only routing and streaming. A UI could not discover which providers were registered or which models an adapter was prepared to recommend. ACP clients therefore received no `model` session config option, so Zed, JetBrains, and VS Code integrations had no model list even though the LLM service already supported runtime switching.
|
||||
|
||||
Model discovery cannot become request validation. The hand-written DeepSeek adapter deliberately forwards arbitrary model ids to a public or private endpoint, while pi-ai has a finite installed catalog that is authoritative for its own request resolution. Treating one shared catalog as a whitelist would remove the private-endpoint behavior that provider routing was designed to preserve.
|
||||
|
||||
@@ -20,7 +20,7 @@ ACP selection must also preserve the provider dimension. The same model id may a
|
||||
|
||||
`LlmAdapter` gains `providerInfo(provider)` and asynchronous `listModels(provider)` methods. Their provider-neutral results are `LlmProviderInfo { id, name }` and `LlmModelInfo { provider, id, name, description? }`. The defaults preserve existing adapter behavior by naming a provider after its route and advertising no models.
|
||||
|
||||
`LlmService.listProviders()` returns detached metadata in registration order. `LlmService.listModels(provider)` delegates to the route owner, validates non-empty ids and names, rejects a mismatched provider or duplicate model id with `INVALID_CATALOG`, and returns detached values. Unknown providers still fail with `NO_ADAPTER`. Provider metadata is validated atomically during `registerAdapter()` so a malformed display record cannot leave a partial registration.
|
||||
`LlmRuntime.listProviders()` returns detached metadata in registration order. `LlmRuntime.listModels(provider)` delegates to the route owner, validates non-empty ids and names, rejects a mismatched provider or duplicate model id with `INVALID_CATALOG`, and returns detached values. Unknown providers still fail with `NO_ADAPTER`. Provider metadata is validated atomically during `registerAdapter()` so a malformed display record cannot leave a partial registration.
|
||||
|
||||
Catalog membership is advisory. It drives selectors and diagnostics but never changes `stream()` routing and never rejects an otherwise valid request. Provider ownership remains exclusive and lifecycle-bound; model ids remain request-time adapter input.
|
||||
|
||||
@@ -28,7 +28,7 @@ Catalog membership is advisory. It drives selectors and diagnostics but never ch
|
||||
|
||||
### Per-session selection in the front end
|
||||
|
||||
A selection is owned by the front end that offers it (today the TUI `/model` selector), never by `LlmService` or `AgentOptions`: those are deployment-wide or creation-wide objects, and mutating them would couple concurrent sessions. Each opaque choice carries the full provider/model pair, because the same model id may appear under multiple routes.
|
||||
A selection is owned by the front end that offers it (today the TUI `/model` selector), never by `LlmRuntime` or `AgentOptions`: those are deployment-wide or creation-wide objects, and mutating them would couple concurrent sessions. Each opaque choice carries the full provider/model pair, because the same model id may appear under multiple routes.
|
||||
|
||||
The ACP automation transport is not a catalog consumer. Its deployment config supplies one optional provider/model target for newly created agents, and it advertises no model selector or configuration-option interface.
|
||||
|
||||
@@ -44,7 +44,7 @@ The request header remains the durable source of truth. When a selection is actu
|
||||
|
||||
**Make catalogs mandatory whitelists.** This conflicts with the hand-written adapter's arbitrary model pass-through and private deployments. The selected adapter already owns authoritative request validation.
|
||||
|
||||
**Store selection in `AgentOptions` or `LlmService`.** Those are creation-wide or deployment-wide objects. Mutating them would couple concurrent sessions and bypass the logged `agent/request` replacement path.
|
||||
**Store selection in `AgentOptions` or `LlmRuntime`.** Those are creation-wide or deployment-wide objects. Mutating them would couple concurrent sessions and bypass the logged `agent/request` replacement path.
|
||||
|
||||
**Persist a new model-selection session event immediately.** An unused UI selection has not affected a model request. Recording the existing request header when the target is consumed preserves the model-visible-if-and-only-if-logged rule without adding a second source of truth.
|
||||
|
||||
|
||||
@@ -8,7 +8,7 @@ Status: implemented
|
||||
|
||||
## 问题
|
||||
|
||||
基于提供方路由的适配器允许每次请求选择 `provider + model`，但 `LlmService` 只暴露路由和流式调用。UI 无法发现已注册的提供方，也无法知道适配器愿意推荐哪些模型。因此，ACP 客户端收不到 `model` 会话配置项；即使 LLM（大语言模型）服务已经支持运行时切换，Zed、JetBrains 和 VS Code 集成仍没有模型列表。
|
||||
基于提供方路由的适配器允许每次请求选择 `provider + model`，但 `LlmRuntime` 只暴露路由和流式调用。UI 无法发现已注册的提供方，也无法知道适配器愿意推荐哪些模型。因此，ACP 客户端收不到 `model` 会话配置项；即使 LLM（大语言模型）服务已经支持运行时切换，Zed、JetBrains 和 VS Code 集成仍没有模型列表。
|
||||
|
||||
模型发现不能变成请求校验。手写 DeepSeek 适配器会把任意模型 ID 原样转发给公开或私有端点，而 pi-ai 的有限安装目录则是其自身请求解析的权威依据。将共享目录视为白名单，会破坏提供方路由需要保留的私有端点能力。
|
||||
|
||||
@@ -20,7 +20,7 @@ ACP 选择还必须保留提供方维度。同一个模型 ID 可能存在于多
|
||||
|
||||
`LlmAdapter` 增加 `providerInfo(provider)` 与异步 `listModels(provider)` 方法。其提供方无关结果分别为 `LlmProviderInfo { id, name }` 和 `LlmModelInfo { provider, id, name, description? }`。默认实现以路由名称作为提供方名称，并且不展示模型，从而保持现有适配器行为。
|
||||
|
||||
`LlmService.listProviders()` 按注册顺序返回元数据副本。`LlmService.listModels(provider)` 委托给路由所有者，校验非空 ID 和名称，并在提供方不匹配或模型 ID 重复时以 `INVALID_CATALOG` 失败，最后返回值的副本。未知提供方仍以 `NO_ADAPTER` 失败。提供方元数据在 `registerAdapter()` 期间进行原子校验，错误展示记录不会留下部分注册。
|
||||
`LlmRuntime.listProviders()` 按注册顺序返回元数据副本。`LlmRuntime.listModels(provider)` 委托给路由所有者，校验非空 ID 和名称，并在提供方不匹配或模型 ID 重复时以 `INVALID_CATALOG` 失败，最后返回值的副本。未知提供方仍以 `NO_ADAPTER` 失败。提供方元数据在 `registerAdapter()` 期间进行原子校验，错误展示记录不会留下部分注册。
|
||||
|
||||
目录成员关系仅提供建议。它驱动选择器与诊断，但不会改变 `stream()` 路由，也不会拒绝原本有效的请求。提供方所有权仍然具有排他性并绑定生命周期；模型 ID 仍是请求时传给适配器的输入。
|
||||
|
||||
@@ -28,7 +28,7 @@ ACP 选择还必须保留提供方维度。同一个模型 ID 可能存在于多
|
||||
|
||||
### 前端内的会话级选择
|
||||
|
||||
选择由提供它的前端拥有（今天是 TUI 的 `/model` 选择器），而不由 `LlmService` 或 `AgentOptions` 拥有：它们是部署级或创建级对象，改动它们会把并发会话耦合在一起。每个不透明选项都携带完整的提供方／模型对，因为同一模型 ID 可能出现在多个路由下。
|
||||
选择由提供它的前端拥有（今天是 TUI 的 `/model` 选择器），而不由 `LlmRuntime` 或 `AgentOptions` 拥有：它们是部署级或创建级对象，改动它们会把并发会话耦合在一起。每个不透明选项都携带完整的提供方／模型对，因为同一模型 ID 可能出现在多个路由下。
|
||||
|
||||
ACP 自动化传输层不是目录消费方。它通过部署配置为新创建的 agent 提供一个可选的提供方／模型目标，不展示模型选择器或配置选项接口。
|
||||
|
||||
@@ -44,7 +44,7 @@ ACP 自动化传输层不是目录消费方。它通过部署配置为新创建
|
||||
|
||||
**将目录设为强制白名单。** 这与手写适配器的任意模型透传和私有部署冲突。请求的权威校验本就属于被选中的适配器。
|
||||
|
||||
**把选择存进 `AgentOptions` 或 `LlmService`。** 它们是创建级或部署级对象。改动它们会把并发会话耦合在一起，并绕过有日志记录的 `agent/request` 替换路径。
|
||||
**把选择存进 `AgentOptions` 或 `LlmRuntime`。** 它们是创建级或部署级对象。改动它们会把并发会话耦合在一起，并绕过有日志记录的 `agent/request` 替换路径。
|
||||
|
||||
**立即持久化一个新的模型选择会话事件。** 未被使用的 UI 选择尚未影响任何模型请求。在目标被消费时记录现有请求头，既保持「模型可见当且仅当有日志」的规则，又不会引入第二个真源。
|
||||
|
||||
|
||||
@@ -2,5 +2,5 @@
|
||||
# side as of the last confirmed-consistent state. Both languages carry equal authority;
|
||||
# after editing either side, bring the other along and re-record with:
|
||||
# pnpm run verify-translation-pairing --write .agents/notes/implemented/architecture/2026-07-15-lsp-capability-seam.md
|
||||
2026-07-15-lsp-capability-seam.md: c407de5275da0a8323b3c6186e178c8b2fafdc31
|
||||
2026-07-15-lsp-capability-seam.zh.md: 7c18eb40fa95c3699c579fd553ac26bfe8c2b13d
|
||||
2026-07-15-lsp-capability-seam.md: 90f9daf4b890bd53621916e492d4d82baaa3e8cc
|
||||
2026-07-15-lsp-capability-seam.zh.md: b85b478b25a375b04e1301d893e53198554641e3
|
||||
|
||||
@@ -17,10 +17,10 @@ Many language servers behave best when the queried document is opened with curre
|
||||
Add LSP as a three-package capability seam with one read-only model tool and one generic local provider implementation:
|
||||
|
||||
1. `@deepseek-ai/dsh-lsp` at `packages/lsp/lsp` owns `ctx.lsp`, provider registration and selection, normalized requests/results, execution control, and structured LSP errors.
|
||||
2. `@deepseek-ai/dsh-lsp-local` at `packages/lsp/lsp-local` adapts configured stdio language servers to the seam. One plugin instance accepts a named server table and registers one isolated provider for each command and extension-to-language-id mapping.
|
||||
2. `@deepseek-ai/dsh-lsp-stdio` at `packages/lsp/lsp-stdio` adapts configured stdio language servers to the seam. One plugin instance accepts a named server table and registers one isolated provider for each command and extension-to-language-id mapping.
|
||||
3. `@deepseek-ai/dsh-tool-lsp` at `packages/lsp/tool-lsp` owns the model-facing `lsp` schema, prompt guidance, argument validation, result limits and formatting, and transport-neutral UI presentation.
|
||||
|
||||
`dsh-lsp-local` is a generic host, not a language-server catalog or installer. Deployments explicitly configure commands and mappings; future presets belong in composition plugins or `cordis.yml` overlays.
|
||||
`dsh-lsp-stdio` is a generic host, not a language-server catalog or installer. Deployments explicitly configure commands and mappings; future presets belong in composition plugins or `cordis.yml` overlays.
|
||||
|
||||
The model and seam expose exactly `goToDefinition`, `findReferences`, `goToImplementation`, and `hover`; no arbitrary JSON-RPC method escapes through `ctx.lsp`. These operation literals match Claude Code's familiar camelCase names while the tool name and `file_path` field remain harness-owned.
|
||||
|
||||
@@ -79,7 +79,7 @@ interface LspService {
|
||||
|
||||
Mapping keys normalize to lowercase, leading-dot extensions selected from `filePath`'s final extension; language ids only synchronize documents. Seam positions and ranges are zero-based UTF-16. `findReferences` always includes declarations: providers enforce this internally, the local mapping sets `context.includeDeclaration: true`, and callers get no flag. Closed result unions normalize navigation to locations and hover to content or `null`; navigation results carry the provider's canonical workspace URI so consumers relativize file URIs in the execution world's namespace. The seam exposes no protocol types, process or document controls, or generic request escape hatch.
|
||||
|
||||
`dsh-lsp-local` owns server configuration, JSON-RPC, process and transient-document state, and protocol translation. It reads through `ctx.fs` and launches through `ctx.subprocess`, depending on their Service Definition packages rather than concrete providers; the [portable execution-world decision](2026-07-28-portable-execution-world-consumers.md) owns that pairing. The server-table key is its provider id. The plugin resolves every server-local setting before registration, rolls back earlier registrations if a later mapping is invalid or conflicts, and retains an independent process pool per provider. `dsh-tool-lsp` runtime-injects only `tools`, `lsp`, and `systemPrompt`, obtains the workspace from `exec.agent?.session.header.cwd` through a package-local `sessionCwd(exec)` helper matching the filesystem tools' lookup, and imports no provider.
|
||||
`dsh-lsp-stdio` owns server configuration, JSON-RPC, process and transient-document state, and protocol translation. It reads through `ctx.fs` and launches through `ctx.subprocess`, depending on their Service Definition packages rather than concrete providers; the [portable execution-world decision](2026-07-28-portable-execution-world-consumers.md) owns that pairing. The server-table key is its provider id. The plugin resolves every server-local setting before registration, rolls back earlier registrations if a later mapping is invalid or conflicts, and retains an independent process pool per provider. `dsh-tool-lsp` runtime-injects only `tools`, `lsp`, and `systemPrompt`, obtains the workspace from `exec.agent?.session.header.cwd` through a package-local `sessionCwd(exec)` helper matching the filesystem tools' lookup, and imports no provider.
|
||||
|
||||
## Model-facing contract
|
||||
|
||||
@@ -104,15 +104,15 @@ The transport-neutral presenter uses `{ card: 'generic', kind: 'search', title,
|
||||
|
||||
## Timeout ownership
|
||||
|
||||
`dsh-tool-lsp` attaches one configurable `timeoutMs` budget, default `60_000`, to the tool definition. `dsh-timeout-policy` enforces it and supplies `exec.signal`, which reaches `ctx.lsp.query`; the budget covers the complete queued open/query/close lifecycle and is not model-configurable.
|
||||
`dsh-tool-lsp` attaches one configurable `timeoutMs` budget, default `60_000`, to the tool definition. `dsh-tool-call-timeout-policy` enforces it and supplies `exec.signal`, which reaches `ctx.lsp.query`; the budget covers the complete queued open/query/close lifecycle and is not model-configurable.
|
||||
|
||||
The seam and provider add no startup or request deadline. Non-tool callers therefore receive no hidden timeout and must supply an `AbortSignal`, using `deadline()` when they need a budget.
|
||||
|
||||
Provider disposal occurs outside tool execution, so `dsh-lsp-local` keeps `shutdownTimeoutMs` (default `5_000`) for `shutdown`/`exit` and `killGraceMs` (default `2_000`) for both request-cancel grace and SIGTERM-to-SIGKILL escalation; the same bounds govern failed-instance cleanup. Timer values above Node's `2_147_483_647` ms scheduling range fail at load. The provider uses `deadline()` and `timeoutOf()` but owns request cancellation, process signals, and awaiting close because timeout notification does not terminate work.
|
||||
Provider disposal occurs outside tool execution, so `dsh-lsp-stdio` keeps `shutdownTimeoutMs` (default `5_000`) for `shutdown`/`exit` and `killGraceMs` (default `2_000`) for both request-cancel grace and SIGTERM-to-SIGKILL escalation; the same bounds govern failed-instance cleanup. Timer values above Node's `2_147_483_647` ms scheduling range fail at load. The provider uses `deadline()` and `timeoutOf()` but owns request cancellation, process signals, and awaiting close because timeout notification does not terminate work.
|
||||
|
||||
## Workspace, filesystem, and document synchronization
|
||||
|
||||
`dsh-lsp-local` canonicalizes and reads through `ctx.fs` in the language server's execution world. It requires the workspace target to be a directory, rejects out-of-workspace sources through provider-owned containment, consumes `streamText`, and enforces `maxDocumentBytes` as chunks arrive; the provider retains regular-file validation and UTF-8 decoding while the protocol consumer owns its document limit. It fuses caller cancellation with provider disposal across each filesystem operation, tracks workspace lookups before they enter a queue, and awaits those lookups during disposal. It does not emit `fs/observed`: only the LSP result is model-visible, so the query does not satisfy read-before-write policy.
|
||||
`dsh-lsp-stdio` canonicalizes and reads through `ctx.fs` in the language server's execution world. It requires the workspace target to be a directory, rejects out-of-workspace sources through provider-owned containment, consumes `streamText`, and enforces `maxDocumentBytes` as chunks arrive; the provider retains regular-file validation and UTF-8 decoding while the protocol consumer owns its document limit. It fuses caller cancellation with provider disposal across each filesystem operation, tracks workspace lookups before they enter a queue, and awaits those lookups during disposal. It does not emit `fs/observed`: only the LSP result is model-visible, so the query does not satisfy read-before-write policy.
|
||||
|
||||
The `read` tool is unsuitable source because its output is windowed, numbered, transcript-visible, and observed. Reading in `tool-lsp` would also assign provider-specific synchronization to the consumer and preclude non-local providers.
|
||||
|
||||
@@ -129,7 +129,7 @@ The canonical workspace target must be a directory. Its target key supplies pool
|
||||
|
||||
## Local server lifecycle and protocol behavior
|
||||
|
||||
`dsh-lsp-local` lazily single-flights one server per `(provider id, canonical workspace target)`. At load it calls `ctx.subprocess.resolveExecutable()` with the configured environment, failing before registration if unavailable; first query launches through raw protocol pipes with no shell and a bounded collected stderr tail. `maxMessageBytes` defaults to `16_000_000`, `maxStderrBytes` to `1_000_000`, and `maxDocumentBytes` to `4_000_000`. A crash fails the active query without replay; a later query may replace the process. Each query starts at most one process, so the MVP has no cross-request restart counter.
|
||||
`dsh-lsp-stdio` lazily single-flights one server per `(provider id, canonical workspace target)`. At load it calls `ctx.subprocess.resolveExecutable()` with the configured environment, failing before registration if unavailable; first query launches through raw protocol pipes with no shell and a bounded collected stderr tail. `maxMessageBytes` defaults to `16_000_000`, `maxStderrBytes` to `1_000_000`, and `maxDocumentBytes` to `4_000_000`. A crash fails the active query without replay; a later query may replace the process. Each query starts at most one process, so the MVP has no cross-request restart counter.
|
||||
|
||||
Initialization uses `processId: null` because the client and server may inhabit different process namespaces. It advertises `general.positionEncodings: ['utf-16']`, `workspace: { workspaceFolders: true, configuration: true }`, `textDocument.hover.contentFormat: ['markdown', 'plaintext']`, and `linkSupport: true` for definition and implementation, with no dynamic registration. Returned operation and synchronization capabilities are authoritative. An omitted server `positionEncoding` defaults to `utf-16`; any other value is a protocol error. Configuration may supply initialization options and `workspace/configuration` responses, but the client rejects `workspace/applyEdit` and never executes commands or edits.
|
||||
|
||||
|
||||
@@ -17,10 +17,10 @@ harness 已具备文本搜索与文件读取能力，但二者都无法识别程
|
||||
将 LSP 建成由三个包组成的能力 seam，其中包含一个只读模型工具和一个通用本地提供方实现：
|
||||
|
||||
1. `packages/lsp/lsp` 下的 `@deepseek-ai/dsh-lsp` 负责 `ctx.lsp`、提供方注册与选择、标准化请求与结果、执行控制，以及结构化 LSP 错误。
|
||||
2. `packages/lsp/lsp-local` 下的 `@deepseek-ai/dsh-lsp-local` 将配置的 stdio 语言服务器适配到该 seam。一个插件实例接收具名服务器表，并为每组命令及扩展名到语言 id 的映射注册一个隔离的提供方。
|
||||
2. `packages/lsp/lsp-stdio` 下的 `@deepseek-ai/dsh-lsp-stdio` 将配置的 stdio 语言服务器适配到该 seam。一个插件实例接收具名服务器表，并为每组命令及扩展名到语言 id 的映射注册一个隔离的提供方。
|
||||
3. `packages/lsp/tool-lsp` 下的 `@deepseek-ai/dsh-tool-lsp` 负责面向模型的 `lsp` schema、提示词指导、参数校验、结果限制与格式化，以及与传输方式无关的 UI 展示。
|
||||
|
||||
`dsh-lsp-local` 是通用 host，不是语言服务器目录或安装器。部署显式配置命令与映射；未来 preset 属于组合插件或 `cordis.yml` overlay。
|
||||
`dsh-lsp-stdio` 是通用 host，不是语言服务器目录或安装器。部署显式配置命令与映射；未来 preset 属于组合插件或 `cordis.yml` overlay。
|
||||
|
||||
模型与 seam 仅公开 `goToDefinition`、`findReferences`、`goToImplementation` 和 `hover`；`ctx.lsp` 不提供任意 JSON-RPC 方法。这些操作字面量与 Claude Code 熟悉的 camelCase 命名一致，而工具名与 `file_path` 字段仍由 harness 自行定义。
|
||||
|
||||
@@ -79,7 +79,7 @@ interface LspService {
|
||||
|
||||
映射键规范化为带前导点的小写扩展名，并按 `filePath` 的最后一个扩展名选择；语言 id 仅用于文档同步。seam 中的位置和范围从零开始按 UTF-16 计数。`findReferences` 始终包含声明：提供方在内部执行该约束，本地映射设置 `context.includeDeclaration: true`，调用方不能配置。封闭结果联合将导航统一为位置，将 `hover` 统一为内容或 `null`；导航结果携带提供方的规范工作区 URI，使消费方在执行世界的命名空间内相对化文件 URI。seam 不公开协议类型、进程或文档控制，也不提供通用请求逃生口。
|
||||
|
||||
`dsh-lsp-local` 负责服务器配置、JSON-RPC、进程与临时文档状态和协议转换。它通过 `ctx.fs` 读取，通过 `ctx.subprocess` 启动，只依赖二者的 Service Definition 包而非具体提供方；[可移植执行环境决策](2026-07-28-portable-execution-world-consumers.md)负责定义这种配对。服务器表的键是提供方 id。插件在注册前解析每个服务器的本地设置；如果后续映射无效或发生冲突，插件会撤销此前的注册，并为每个提供方保留独立进程池。`dsh-tool-lsp` 在运行时只注入 `tools`、`lsp` 和 `systemPrompt`，通过包内的 `sessionCwd(exec)` 辅助函数从 `exec.agent?.session.header.cwd` 取得工作区，其取值方式与文件系统工具一致，也不导入提供方。
|
||||
`dsh-lsp-stdio` 负责服务器配置、JSON-RPC、进程与临时文档状态和协议转换。它通过 `ctx.fs` 读取，通过 `ctx.subprocess` 启动，只依赖二者的 Service Definition 包而非具体提供方；[可移植执行环境决策](2026-07-28-portable-execution-world-consumers.md)负责定义这种配对。服务器表的键是提供方 id。插件在注册前解析每个服务器的本地设置；如果后续映射无效或发生冲突，插件会撤销此前的注册，并为每个提供方保留独立进程池。`dsh-tool-lsp` 在运行时只注入 `tools`、`lsp` 和 `systemPrompt`，通过包内的 `sessionCwd(exec)` 辅助函数从 `exec.agent?.session.header.cwd` 取得工作区，其取值方式与文件系统工具一致，也不导入提供方。
|
||||
|
||||
## 面向模型的约定
|
||||
|
||||
@@ -104,15 +104,15 @@ interface LspToolInput {
|
||||
|
||||
## 超时归属
|
||||
|
||||
`dsh-tool-lsp` 将一个可配置的 `timeoutMs` 预算附加到工具定义，默认值为 `60_000`。`dsh-timeout-policy` 执行预算并提供传入 `ctx.lsp.query` 的 `exec.signal`；该预算覆盖排队、打开、查询和关闭的完整生命周期，模型不可配置。
|
||||
`dsh-tool-lsp` 将一个可配置的 `timeoutMs` 预算附加到工具定义，默认值为 `60_000`。`dsh-tool-call-timeout-policy` 执行预算并提供传入 `ctx.lsp.query` 的 `exec.signal`；该预算覆盖排队、打开、查询和关闭的完整生命周期，模型不可配置。
|
||||
|
||||
seam 和提供方不增加启动或请求截止时间。非工具调用方不会获得隐藏超时，必须自行提供 `AbortSignal`，并在需要预算时使用 `deadline()`。
|
||||
|
||||
提供方 dispose 发生在工具执行之外，因此 `dsh-lsp-local` 保留 `shutdownTimeoutMs`（默认 `5_000`）限制 `shutdown`/`exit`，以及 `killGraceMs`（默认 `2_000`），同时用于限制请求取消宽限期和从 SIGTERM 升级到 SIGKILL 的宽限期；失败实例的清理也使用相同边界。定时器值超过 Node `2_147_483_647` ms 的调度范围时，插件加载失败。提供方使用 `deadline()` 和 `timeoutOf()`，但仍负责请求取消、进程信号和等待关闭，因为超时通知不会终止工作。
|
||||
提供方 dispose 发生在工具执行之外，因此 `dsh-lsp-stdio` 保留 `shutdownTimeoutMs`（默认 `5_000`）限制 `shutdown`/`exit`，以及 `killGraceMs`（默认 `2_000`），同时用于限制请求取消宽限期和从 SIGTERM 升级到 SIGKILL 的宽限期；失败实例的清理也使用相同边界。定时器值超过 Node `2_147_483_647` ms 的调度范围时，插件加载失败。提供方使用 `deadline()` 和 `timeoutOf()`，但仍负责请求取消、进程信号和等待关闭，因为超时通知不会终止工作。
|
||||
|
||||
## 工作区、文件系统与文档同步
|
||||
|
||||
`dsh-lsp-local` 在语言服务器的执行环境中通过 `ctx.fs` 规范化并读取文件。它要求工作区目标是目录，使用提供方自有的 containment 拒绝工作区外的源文件，消费 `streamText`，并在分片到达时执行 `maxDocumentBytes` 上限；普通文件校验和 UTF-8 解码仍由提供方负责，文档上限则由协议消费方负责。它会针对每项文件系统操作合并调用方取消与提供方 dispose，跟踪尚未进入队列的工作区查找，并在 dispose 期间等待这些查找结算。它不发送 `fs/observed`：只有 LSP 结果对模型可见，因此查询不满足写前读取策略。
|
||||
`dsh-lsp-stdio` 在语言服务器的执行环境中通过 `ctx.fs` 规范化并读取文件。它要求工作区目标是目录，使用提供方自有的 containment 拒绝工作区外的源文件，消费 `streamText`，并在分片到达时执行 `maxDocumentBytes` 上限；普通文件校验和 UTF-8 解码仍由提供方负责，文档上限则由协议消费方负责。它会针对每项文件系统操作合并调用方取消与提供方 dispose，跟踪尚未进入队列的工作区查找，并在 dispose 期间等待这些查找结算。它不发送 `fs/observed`：只有 LSP 结果对模型可见，因此查询不满足写前读取策略。
|
||||
|
||||
`read` 工具的输出带窗口与行号，进入 transcript（文本记录）且已被观察，不适合作为源文件。在 `tool-lsp` 内读取还会把提供方专用同步职责交给消费方，并排除非本地提供方。
|
||||
|
||||
@@ -129,7 +129,7 @@ seam 和提供方不增加启动或请求截止时间。非工具调用方不会
|
||||
|
||||
## 本地服务器生命周期与协议行为
|
||||
|
||||
`dsh-lsp-local` 按 `(provider id, canonical workspace target)` 懒启动一个服务器，并通过 single-flight 合并启动。插件加载时，它使用已配置的环境调用 `ctx.subprocess.resolveExecutable()`；命令不可用时在注册前失败。首次查询通过原始协议管道启动服务器，不经过 shell，并收集有界的 stderr 尾部。`maxMessageBytes` 默认值为 `16_000_000`，`maxStderrBytes` 默认值为 `1_000_000`，`maxDocumentBytes` 默认值为 `4_000_000`。崩溃使当前查询失败且不重放；后续查询可以替换进程。每次查询最多启动一个进程，因此 MVP 不设置跨请求重启计数器。
|
||||
`dsh-lsp-stdio` 按 `(provider id, canonical workspace target)` 懒启动一个服务器，并通过 single-flight 合并启动。插件加载时，它使用已配置的环境调用 `ctx.subprocess.resolveExecutable()`；命令不可用时在注册前失败。首次查询通过原始协议管道启动服务器，不经过 shell，并收集有界的 stderr 尾部。`maxMessageBytes` 默认值为 `16_000_000`，`maxStderrBytes` 默认值为 `1_000_000`，`maxDocumentBytes` 默认值为 `4_000_000`。崩溃使当前查询失败且不重放；后续查询可以替换进程。每次查询最多启动一个进程，因此 MVP 不设置跨请求重启计数器。
|
||||
|
||||
初始化使用 `processId: null`，因为客户端与服务器可能位于不同的进程命名空间。它声明 `general.positionEncodings: ['utf-16']`、`workspace: { workspaceFolders: true, configuration: true }`、`textDocument.hover.contentFormat: ['markdown', 'plaintext']`，以及 definition 与 implementation 的 `linkSupport: true`，但不支持动态注册。服务器返回的操作与同步能力均为真源。服务器省略 `positionEncoding` 时默认为 `utf-16`；其他值均属于协议错误。配置可以提供初始化选项和 `workspace/configuration` 响应，但客户端拒绝 `workspace/applyEdit`，绝不执行命令或编辑。
|
||||
|
||||
|
||||
@@ -2,5 +2,5 @@
|
||||
# side as of the last confirmed-consistent state. Both languages carry equal authority;
|
||||
# after editing either side, bring the other along and re-record with:
|
||||
# pnpm run verify-translation-pairing --write .agents/notes/implemented/architecture/2026-07-15-replay-token-meter-service.md
|
||||
2026-07-15-replay-token-meter-service.md: 4013cc92f67597a9b87edfc97d14c7f47f0523ac
|
||||
2026-07-15-replay-token-meter-service.zh.md: 82e6175f6c6192c6410a50af199e2c1e62176b5e
|
||||
2026-07-15-replay-token-meter-service.md: 10261189fca621b305b3ec6347b54471d81c8ba3
|
||||
2026-07-15-replay-token-meter-service.zh.md: 4e460b3c32630f205f447a16205bd4c9b1d89faf
|
||||
|
||||
@@ -6,7 +6,7 @@ English | [中文](2026-07-15-replay-token-meter-service.zh.md)
|
||||
|
||||
## Problem
|
||||
|
||||
Context pressure is useful outside compaction. A compaction backend, an overflow guard, or a future request-policy plugin can all need the same answer: how many tokens does the durable request consume? Keeping that fold inside `dsh-compact-basic` duplicates replay logic, makes measurement unavailable without compaction, and encourages callers to reuse stale accounting.
|
||||
Context pressure is useful outside compaction. A compaction backend, an overflow guard, or a future request-policy plugin can all need the same answer: how many tokens does the durable request consume? Keeping that fold inside `dsh-compaction-basic` duplicates replay logic, makes measurement unavailable without compaction, and encourages callers to reuse stale accounting.
|
||||
|
||||
Provider usage is not a complete answer. It describes one successful call under one exact request envelope, while the current surface can grow, shrink, or be replaced afterward. Sessions also switch providers and models, old logs can omit the chunk seqs behind an assistant message, and usage fields separate input, cache-read, cache-write, output, and reasoning counts. A useful service therefore combines the latest exact anchor with conservative heuristic repricing and exposes the log revision consumed by each result.
|
||||
|
||||
@@ -14,7 +14,7 @@ Provider usage is not a complete answer. It describes one successful call under
|
||||
|
||||
### One concrete LLM-family service
|
||||
|
||||
`@deepseek-ai/dsh-token-meter` is one concrete package under `packages/llm/` and registers `ctx.tokenMeter`. It is not split into an interface and backend before a second implementation exists. `TokenMeterService` itself exposes `measure(session, requestHeader?)` and `estimateMessage(message)`; consumers call the singleton service directly.
|
||||
`@deepseek-ai/dsh-token-meter` is one concrete package under `packages/llm/` and registers `ctx.tokenMeter`. It is not split into an interface and backend before a second implementation exists. `TokenMeter` itself exposes `measure(session, requestHeader?)` and `estimateMessage(message)`; consumers call the singleton service directly.
|
||||
|
||||
The service has no configuration. Estimation uses a fixed four-characters-per-token heuristic plus structural overhead. There are no model profiles, capacity settings, density settings, tokenizer backends, or language-specific strategies. Exact provider/model capacity is a separate adapter-owned query, as specified by the [routed model context and compaction policy Agent Note](2026-07-20-routed-model-context-and-compaction-policy.md).
|
||||
|
||||
@@ -30,9 +30,9 @@ Usage sums the disjoint input, cache-read, cache-write, and output buckets. Reas
|
||||
|
||||
### Compact-basic consumes, but does not own, measurement
|
||||
|
||||
`dsh-compact-basic` requires `ctx.tokenMeter`; `CompactService` gains no token methods or types. Configuration, the region transaction, and summarization stay in separate modules; the service registers automatic listeners itself, while `summarize()` remains its sole subclass hook. The singleton meter consistently prices pressure, retention, shadowed content, cited source events, and non-shrinking-summary rejection.
|
||||
`dsh-compaction-basic` requires `ctx.tokenMeter`; `CompactionEngine` gains no token methods or types. Configuration, the region transaction, and summarization stay in separate modules; the service registers automatic listeners itself, while `summarize()` remains its sole subclass hook. The singleton meter consistently prices pressure, retention, shadowed content, cited source events, and non-shrinking-summary rejection.
|
||||
|
||||
Automatic compaction uses one unified measurement for each threshold-and-retention decision. The region transaction measures after appending its durable `compact/start` lock and again after asynchronous summarization, then compares the detached surface-node vectors. An intervening surface mutation prevents replacement; `logRevision` may advance for unrelated log-only facts without invalidating an unchanged selected span.
|
||||
Automatic compaction uses one unified measurement for each threshold-and-retention decision. The region transaction measures after appending its durable `compaction/start` lock and again after asynchronous summarization, then compares the detached surface-node vectors. An intervening surface mutation prevents replacement; `logRevision` may advance for unrelated log-only facts without invalidating an unchanged selected span.
|
||||
|
||||
Compact policy has service-wide defaults: threshold ratio `0.8`, retained-tail ratio `0.16`, `summarizationProvider: ''`, `summarizationModel: ''`, `maxTokens: 8192`, `compactionRetries: 1`, `maxOverflowRetries: 1`, and `auto: true`. Top-level fields apply to every routed target; exact provider/model entries in `modelPolicies` partially override them. Pressure scales ratios against capacity resolved from the owning adapter, and `retainTokens` may replace `retainRatio`; retention must remain below the resulting threshold. The summarization provider and model must both be set or both be empty; an empty pair resolves the latest logged request target, then the `AgentOptions` pair.
|
||||
|
||||
@@ -40,20 +40,20 @@ Automatic pressure runs at `agent/pre-step` before request derivation and measur
|
||||
|
||||
## Testing
|
||||
|
||||
Unit tests cover fixed estimation, envelope invalidation and anchor replacement, replay boundaries, immutable snapshots, routed pressure, convergence, overflow generation proof, and rollback. A real Loader/Include fixture verifies the zero-config token-meter and compact-basic load path in dependency order.
|
||||
Unit tests cover fixed estimation, envelope invalidation and anchor replacement, replay boundaries, immutable snapshots, routed pressure, convergence, overflow generation proof, and rollback. A real Loader/Include fixture verifies the zero-config token-meter and compaction-basic load path in dependency order.
|
||||
|
||||
## Alternatives considered
|
||||
|
||||
- **Keep estimation inside `CompactService`** — rejected because measurement has consumers and replay semantics independent of compaction; it would also force every compactor to expose the same unrelated API.
|
||||
- **Keep estimation inside `CompactionEngine`** — rejected because measurement has consumers and replay semantics independent of compaction; it would also force every compactor to expose the same unrelated API.
|
||||
- **Split a token-meter interface from a heuristic backend immediately** — rejected because only one implementation exists. One concrete service preserves the future seam without speculative packages or configuration.
|
||||
- **Put model-keyed windows and density profiles in the meter** — rejected because replay estimation does not own model routing or capacity facts. The route-owning adapter exposes capacity, while compact-basic owns the consumer-specific threshold and retention policy.
|
||||
- **Put model-keyed windows and density profiles in the meter** — rejected because replay estimation does not own model routing or capacity facts. The route-owning adapter exposes capacity, while compaction-basic owns the consumer-specific threshold and retention policy.
|
||||
- **Keep separate scalar and surface measurements** — rejected because callers would need two reads and revision matching for one decision. A scalar-only read could avoid cloning nodes below threshold, but the split API introduces a caller-side race window; the unified snapshot accepts O(surface) cloning in exchange for coherence.
|
||||
- **Treat provider usage as portable between envelopes** — rejected because model, tools, prefixes, and call config are request facts. Mismatch reprices the whole current request.
|
||||
|
||||
## Consequences
|
||||
|
||||
- Token pressure has one replay-aware owner that compaction and future plugins can share.
|
||||
- The default makes the meter a zero-config composition entry; deployments configure capacity on each route-owning adapter and optional policy overrides on compact-basic.
|
||||
- The default makes the meter a zero-config composition entry; deployments configure capacity on each route-owning adapter and optional policy overrides on compaction-basic.
|
||||
- Fixed heuristic pricing remains an estimate of provider behavior and is not an exact tokenizer or request serializer.
|
||||
- Every measurement clones the current positional surface and therefore costs O(surface), including pressure checks that finish below threshold.
|
||||
- Measurements fail loudly on malformed durable boundaries. This turns corrupted replay into a named integration failure instead of silently drifting pressure.
|
||||
|
||||
@@ -6,7 +6,7 @@ Status: implemented
|
||||
|
||||
## 问题
|
||||
|
||||
上下文压力并不只对压缩（compaction）有用。压缩后端、溢出保护或未来的请求策略插件都可能需要回答同一个问题：持久请求消耗了多少 token？如果把该折叠逻辑留在 `dsh-compact-basic` 内部，就会重复实现回放逻辑，使未加载压缩的调用方无法使用计量，并诱使调用方复用陈旧的核算结果。
|
||||
上下文压力并不只对压缩（compaction）有用。压缩后端、溢出保护或未来的请求策略插件都可能需要回答同一个问题：持久请求消耗了多少 token？如果把该折叠逻辑留在 `dsh-compaction-basic` 内部，就会重复实现回放逻辑，使未加载压缩的调用方无法使用计量，并诱使调用方复用陈旧的核算结果。
|
||||
|
||||
提供方 usage 也不是完整答案。它只描述某个精确请求信封下的一次成功调用，而当前表层之后还可能增长、缩小或被替换。会话也可能切换提供方与模型，旧日志可能缺少构成 assistant 消息的分片 seq，usage 字段还会分别报告输入、缓存读取、缓存写入、输出与推理计数。因此，可用的服务必须把最新精确锚点与保守的启发式重新定价结合起来，并公开每个结果已经消费的日志修订号。
|
||||
|
||||
@@ -14,7 +14,7 @@ Status: implemented
|
||||
|
||||
### 一个具体的 LLM（大语言模型）家族服务
|
||||
|
||||
`@deepseek-ai/dsh-token-meter` 是 `packages/llm/` 下的单个具体包，并注册 `ctx.tokenMeter`。在第二种实现出现之前，它不会被拆成接口与后端。`TokenMeterService` 本身公开 `measure(session, requestHeader?)` 与 `estimateMessage(message)`；消费方直接调用这个单例服务。
|
||||
`@deepseek-ai/dsh-token-meter` 是 `packages/llm/` 下的单个具体包，并注册 `ctx.tokenMeter`。在第二种实现出现之前，它不会被拆成接口与后端。`TokenMeter` 本身公开 `measure(session, requestHeader?)` 与 `estimateMessage(message)`；消费方直接调用这个单例服务。
|
||||
|
||||
服务没有配置。估算采用固定的每 token 四个字符启发式规则，并加上结构开销。服务不提供模型 profile、容量设置、密度设置、分词器后端或语言专用策略。对精确提供方/模型容量的查询由适配器单独负责，具体见[路由模型上下文与压缩策略 Agent Note](2026-07-20-routed-model-context-and-compaction-policy.md)。
|
||||
|
||||
@@ -28,11 +28,11 @@ Status: implemented
|
||||
|
||||
Usage 会对互不重叠的输入、缓存读取、缓存写入与输出 bucket 求和，不会再次加入推理计数。每次成功模型调用都会记录 `assistant/message`，包括无内容调用与达到 token 上限的调用，并带上精确的更早分片 seq。显式的空 `sourceEventSeqs` 列表表示已知为空的提供方流；旧日志中缺失的列表则保守地把持久 assistant 输出视为提供方输出。
|
||||
|
||||
### compact-basic 消费计量，但不拥有计量
|
||||
### compaction-basic 消费计量，但不拥有计量
|
||||
|
||||
`dsh-compact-basic` 要求 `ctx.tokenMeter`；`CompactService` 不增加 token 方法或类型。配置、区域事务与摘要分别保留在独立模块中，服务自身注册自动监听器，而 `summarize()` 仍是唯一的子类钩子。单例计量器一致用于压力、保留、被遮蔽内容、引用的源事件以及非缩小摘要拒绝的定价。
|
||||
`dsh-compaction-basic` 要求 `ctx.tokenMeter`；`CompactionEngine` 不增加 token 方法或类型。配置、区域事务与摘要分别保留在独立模块中，服务自身注册自动监听器，而 `summarize()` 仍是唯一的子类钩子。单例计量器一致用于压力、保留、被遮蔽内容、引用的源事件以及非缩小摘要拒绝的定价。
|
||||
|
||||
自动压缩的每次阈值与保留联合决策只使用一次统一计量。区域事务会在追加持久 `compact/start` 锁后执行计量，在异步摘要完成后再次计量，随后比较分离的表层节点向量。期间发生的表层变更会阻止替换；`logRevision` 可以因无关的纯日志事实而推进，而不会使未变的选定范围失效。
|
||||
自动压缩的每次阈值与保留联合决策只使用一次统一计量。区域事务会在追加持久 `compaction/start` 锁后执行计量，在异步摘要完成后再次计量，随后比较分离的表层节点向量。期间发生的表层变更会阻止替换；`logRevision` 可以因无关的纯日志事实而推进，而不会使未变的选定范围失效。
|
||||
|
||||
压缩策略采用服务级默认值：阈值比例 `0.8`、保留尾部比例 `0.16`、`summarizationProvider: ''`、`summarizationModel: ''`、`maxTokens: 8192`、`compactionRetries: 1`、`maxOverflowRetries: 1` 与 `auto: true`。顶层字段适用于每个路由目标；`modelPolicies` 中的精确提供方/模型项可以部分覆盖这些字段。压力检查以所属适配器解析的容量为基准换算这些比例，`retainTokens` 可以替代 `retainRatio`；保留值必须小于最终阈值。摘要提供方与模型必须同时设置或同时为空；空组合先解析最近记录的请求目标，再使用 `AgentOptions` 中的组合。
|
||||
|
||||
@@ -40,20 +40,20 @@ Usage 会对互不重叠的输入、缓存读取、缓存写入与输出 bucket
|
||||
|
||||
## 测试
|
||||
|
||||
单元测试覆盖固定估算、信封失效与锚点替换、回放边界、不可变快照、已路由压力、收敛、溢出 generation 证明与回滚。真实 Loader/Include fixture（测试前置数据）验证零配置 token-meter 与 compact-basic 按依赖顺序加载的路径。
|
||||
单元测试覆盖固定估算、信封失效与锚点替换、回放边界、不可变快照、已路由压力、收敛、溢出 generation 证明与回滚。真实 Loader/Include fixture（测试前置数据）验证零配置 token-meter 与 compaction-basic 按依赖顺序加载的路径。
|
||||
|
||||
## 考虑过的替代方案
|
||||
|
||||
- **把估算保留在 `CompactService` 内**——不予采纳，因为计量拥有独立于压缩的消费方与回放语义；它还会强迫每个压缩器暴露同一套无关 API。
|
||||
- **把估算保留在 `CompactionEngine` 内**——不予采纳，因为计量拥有独立于压缩的消费方与回放语义；它还会强迫每个压缩器暴露同一套无关 API。
|
||||
- **立即把 token meter 拆成接口与启发式后端**——不予采纳，因为目前只有一种实现。单个具体服务保留未来的 seam，同时避免推测性的包与配置。
|
||||
- **把模型键控窗口与密度 profile 放进 meter**——不予采纳，因为回放估算不拥有模型路由或容量事实。路由所属适配器公开容量，compact-basic 则拥有消费方专用的阈值与保留策略。
|
||||
- **把模型键控窗口与密度 profile 放进 meter**——不予采纳，因为回放估算不拥有模型路由或容量事实。路由所属适配器公开容量，compaction-basic 则拥有消费方专用的阈值与保留策略。
|
||||
- **保留独立的标量与表层计量**——不予采纳，因为消费方必须为一次决策执行两次读取并匹配修订号。仅读取标量可以避免在低于阈值时复制节点，但拆分 API 会在消费方引入竞态窗口；统一快照接受 O(surface) 复制成本，以换取结果一致性。
|
||||
- **在不同信封之间移用提供方 usage**——不予采纳，因为模型、工具、前缀与调用配置都是请求事实。不匹配时会重新定价完整当前请求。
|
||||
|
||||
## 后果
|
||||
|
||||
- Token 压力拥有一个可供压缩与未来插件共享的回放感知所有者。
|
||||
- 默认值让 meter 成为零配置组合项；部署时在各个路由所属适配器上配置容量，并在 compact-basic 上配置可选策略覆盖。
|
||||
- 默认值让 meter 成为零配置组合项；部署时在各个路由所属适配器上配置容量，并在 compaction-basic 上配置可选策略覆盖。
|
||||
- 固定启发式定价仍然只是提供方行为的估计，并不是精确分词器或请求序列化器。
|
||||
- 每次计量都会复制当前带位置信息的表层，因此成本为 O(surface)，低于阈值即可结束的压力检查也不例外。
|
||||
- 遇到畸形持久边界时，计量会明确失败。这会把损坏的回放转化为具名集成错误，而不是让压力静默漂移。
|
||||
|
||||
@@ -2,5 +2,5 @@
|
||||
# side as of the last confirmed-consistent state. Both languages carry equal authority;
|
||||
# after editing either side, bring the other along and re-record with:
|
||||
# pnpm run verify-translation-pairing --write .agents/notes/implemented/architecture/2026-07-19-gui-layering-and-rpc-protocol.md
|
||||
2026-07-19-gui-layering-and-rpc-protocol.md: deba5e81c35e1572e2e6dc68b48234414ac9a7d5
|
||||
2026-07-19-gui-layering-and-rpc-protocol.zh.md: f450f04b48c016643bf498b9ea88f2bfde7d79ea
|
||||
2026-07-19-gui-layering-and-rpc-protocol.md: 22bebb0951ca3ffa0a0679396c4c54ef32eca279
|
||||
2026-07-19-gui-layering-and-rpc-protocol.zh.md: c38328cc2e692e8eed44a3dbf4149207820265c9
|
||||
|
||||
@@ -30,8 +30,8 @@ Directories layer as follows:
|
||||
- **Static-arrival entry packages** (`connection`, `runtime`, `ui-theme`, `i18n`, `hmr`): no `dsh.client` key and no browser bundle — the shell bundles their `src/client/` half and registers it with `ctx.modules`; they are governed as entries of the host-authored graph like everything else.
|
||||
- **Fetch-arrival plugin packages** (`ui-layout`, `ui-sidebar`, `ui-conversation`, `ui-trajectory`): dual-entry — the root index is the node half (an empty `apply`, existing so the host Loader governs lifecycle and the web plugin registry discovers the package.json `dsh.client` declaration); the implementation lives under `src/client/`, shipped as the `./client` subpath (a tsdown closure-factory bundle). Cross-plugin consumption of `/client` is type-only; value cooperation goes through cordis services.
|
||||
- `apps/` holds the externally exported applications, assembled from Client / Host mixtures.
|
||||
- `apps/web` (`dsh-frontend`) is the vite application: a thin `main.ts` over the shell API exported by `dsh-client-web`.
|
||||
- `apps/cli` (`@deepseek-ai/dsh`) dispatches commands: `dsh web` = Host + webserver + the built `dsh-frontend` dist; `dsh --profile headless` = [a direct core Agent/Session entry point](2026-08-09-headless-direct-core-entry-point.md), with zero Host, HTTP, or browser layer.
|
||||
- `apps/web` (`dsh-web-frontend`) is the vite application: a thin `main.ts` over the shell API exported by `dsh-client-web`.
|
||||
- `apps/cli` (`@deepseek-ai/dsh`) dispatches commands: `dsh web` = Host + webserver + the built `dsh-web-frontend` dist; `dsh --profile headless` = [a direct core Agent/Session entry point](2026-08-09-headless-direct-core-entry-point.md), with zero Host, HTTP, or browser layer.
|
||||
- A future Electron application reuses the same web client packages over an IPC fetch carrier.
|
||||
|
||||
```
|
||||
@@ -67,7 +67,7 @@ On the protocol side: TS interfaces (`packages/host/apiproxy/src/api/`, zero Nod
|
||||
| Carrier layer | `dsh-host-webserver` | Web HTTP and upgrade: static serving + `/api/*`→handler forwarding + WebSocket upgrade route + close semantics; plugin bundle endpoint + `__DSH_BOOT__` manifest injection (fed by the web plugin registry) | Web (browser access) only; zero workspace dependencies (the registry arrives by structural injection); Electron does not reuse it |
|
||||
| Client libraries | `dsh-client-ui-slots` / `dsh-client-web-react` / `dsh-client-ui-primitives` | Slot registry core / ctx↔React glue / pure React atoms | Zero cordis runtime dependency in components; seeded into the loader module table by the shell |
|
||||
| Client plugins | `dsh-client-connection` / `dsh-client-runtime` / `dsh-client-ui-theme` / `dsh-client-i18n` / `dsh-client-ui-layout` / `dsh-client-ui-sidebar` / `dsh-client-ui-conversation` / `dsh-client-ui-trajectory` | Browser-side cordis plugin tree (wire consumer, core services, theme, i18n, layout, sidebar, conversation, trajectory) — see the web client architecture note | Dual entry (node half = empty apply; implementation in `src/client/`); the consumption face goes exclusively through ApiProxy |
|
||||
| Application | `@deepseek-ai/dsh` (apps/cli) + `dsh-frontend` (apps/web, the vite application) | Coarse bin dispatch + one assembly module per application (web.ts / headless.ts); the vite app is a thin main over the `dsh-client-web` shell surface | Applications use dynamic imports so they never load each other; workspace knowledge like dist location stays in the app |
|
||||
| Application | `@deepseek-ai/dsh` (apps/cli) + `dsh-web-frontend` (apps/web, the vite application) | Coarse bin dispatch + one assembly module per application (web.ts / headless.ts); the vite app is a thin main over the `dsh-client-web` shell surface | Applications use dynamic imports so they never load each other; workspace knowledge like dist location stays in the app |
|
||||
|
||||
#### Naming rule
|
||||
|
||||
@@ -245,7 +245,7 @@ Every client consumes one contract: adding a unary method is a five-step mechani
|
||||
| Consuming clients connecting to ctx directly (skipping the apiproxy layer) | Clients require wire validation, observability, and multi-client consistency. Direct headless is a local entry point with no client boundary and uses the public Agent/Session seams rather than a client command plane |
|
||||
| webserver depending on runtime (saving the handler injection) | Structural-typing injection keeps webserver reusable by sidecars/tests with zero workspace deps; a package dependency would drag assembly knowledge into the carrier layer |
|
||||
| Package names without the group prefix (continuing dsh-<tail>) | `dsh-runtime`/`dsh-web-ui` lose their belonging in the flat npm namespace; the cost is one explicit paths entry per package |
|
||||
| Reusing the in-repo JSON-RPC 2.0 (dsh-jsonrpc) | Numeric error codes degrade to a single fallback code, contracts get aligned by hand in two copies, and naming drifts without a convention |
|
||||
| Reusing the in-repo JSON-RPC 2.0 (dsh-sdk-jsonrpc-server) | Numeric error codes degrade to a single fallback code, contracts get aligned by hand in two copies, and naming drifts without a convention |
|
||||
| A three-envelope model (Request/Response/Frame envelopes, signatures direction-blind) | rpcId correlation is logical-layer; frame and response direction semantics inferred from the channel break the moment the carrier changes |
|
||||
| Named Request/Response type pairs as the source of truth (map registering type pairs) | Flat named types are a second name for the same fact; signature inference makes adding a method a one-place change |
|
||||
| REST-style paths | The consumer is our own client with no third-party REST expectations; RPC mapping straight onto the method table is more mechanical |
|
||||
|
||||
@@ -28,8 +28,8 @@ Status: implemented
|
||||
- **静态到达 entry 包**（`connection`、`runtime`、`ui-theme`、`i18n`、`hmr`）：无 `dsh.client` 键、无浏览器 bundle——壳把它们的 `src/client/` 半边打进自己的 bundle 并向 `ctx.modules` 登记；它们与其余单元一样，作为 host 独家撰写的图里的 entry 受治理。
|
||||
- **fetch 到达插件包**（`ui-layout`、`ui-sidebar`、`ui-conversation`、`ui-trajectory`）：双入口——根入口是 node 半边（空 `apply`，其存在是为了让 host Loader 管辖生命周期、让 web 插件注册表发现 package.json 的 `dsh.client` 声明）；实现住在 `src/client/` 下，经 `./client` 子路径发布（tsdown 闭包工厂 bundle）。跨插件消费 `/client` 只限类型；值层面的协作走 cordis 服务。
|
||||
- `apps/` 作为对外导出的应用入口，可以由 Client / Host 混合组装。
|
||||
- `apps/web`（`dsh-frontend`）是 vite 应用：`dsh-client-web` 导出的壳 API 之上的一层薄 `main.ts`。
|
||||
- `apps/cli`（`@deepseek-ai/dsh`）分发命令：`dsh web` = Host + webserver + 构建出的 `dsh-frontend` dist；`dsh --profile headless` = [直接使用核心 Agent／Session 的入口](2026-08-09-headless-direct-core-entry-point.md)，不含 Host、HTTP 或浏览器层。
|
||||
- `apps/web`（`dsh-web-frontend`）是 vite 应用：`dsh-client-web` 导出的壳 API 之上的一层薄 `main.ts`。
|
||||
- `apps/cli`（`@deepseek-ai/dsh`）分发命令：`dsh web` = Host + webserver + 构建出的 `dsh-web-frontend` dist；`dsh --profile headless` = [直接使用核心 Agent／Session 的入口](2026-08-09-headless-direct-core-entry-point.md)，不含 Host、HTTP 或浏览器层。
|
||||
- 将来的 Electron 应用经由 IPC fetch 载体复用同一套 web client 包。
|
||||
|
||||
```
|
||||
@@ -65,7 +65,7 @@ TypeScript 以 solution 根引用的**两个聚合 program** 检查（`tsconfig.
|
||||
| 承载层 | `dsh-host-webserver` | Web HTTP 与 upgrade：静态服务 + `/api/*`→handler 转发 + WebSocket upgrade route + close 语义；插件 bundle 端点 + `__DSH_BOOT__` manifest（元数据清单）注入（由 web 插件注册表供给） | Web（浏览器访问）专用；零 workspace 依赖（注册表经结构注入到达）；Electron 不复用它 |
|
||||
| client 库 | `dsh-client-ui-slots` / `dsh-client-web-react` / `dsh-client-ui-primitives` | slot 注册表核心 / ctx↔React 胶合 / 纯 React 原子组件 | 组件零 cordis 运行时依赖；由壳播种进 loader 模块表 |
|
||||
| client 插件 | `dsh-client-connection` / `dsh-client-runtime` / `dsh-client-ui-theme` / `dsh-client-i18n` / `dsh-client-ui-layout` / `dsh-client-ui-sidebar` / `dsh-client-ui-conversation` / `dsh-client-ui-trajectory` | 浏览器侧 cordis 插件树（wire 消费方、核心服务、主题、i18n、布局、侧栏、对话、轨迹）——见 Web 客户端架构笔记 | 双入口（node 半边=空 apply；实现在 `src/client/`）；消费面唯一经 ApiProxy |
|
||||
| 应用 | `@deepseek-ai/dsh`（apps/cli）+ `dsh-frontend`（apps/web，vite 应用） | bin 粗分发 + 每个应用一个拼装模块（web.ts / headless.ts）；vite 应用是 `dsh-client-web` 壳表面之上的薄 main | 各应用使用动态 import，因此不会互相加载；dist 定位等 workspace 知识留在 app |
|
||||
| 应用 | `@deepseek-ai/dsh`（apps/cli）+ `dsh-web-frontend`（apps/web，vite 应用） | bin 粗分发 + 每个应用一个拼装模块（web.ts / headless.ts）；vite 应用是 `dsh-client-web` 壳表面之上的薄 main | 各应用使用动态 import，因此不会互相加载；dist 定位等 workspace 知识留在 app |
|
||||
|
||||
#### 命名规则
|
||||
|
||||
@@ -243,7 +243,7 @@ export type ResponseValue<K> =
|
||||
| 消费型 client 直连 ctx（省 apiproxy 一层） | client 需要 wire 校验、观测与多 client 一致性。直接 headless 是没有 client 边界的本地入口，使用公开的 Agent／Session seam，而不是 client 命令面 |
|
||||
| webserver 依赖 runtime（省 handler 注入） | 结构 typing 注入让 webserver 可被 sidecar/测试复用且零 workspace 依赖；包依赖会把装配知识拖进承载层 |
|
||||
| 包名不带组前缀（沿用 dsh-<尾段>） | `dsh-runtime`/`dsh-web-ui` 在扁平 npm 命名空间里失去归属信息；代价只是每包一条显式 paths |
|
||||
| 复用仓内 JSON-RPC 2.0（dsh-jsonrpc） | 数字错误码退化成单码兜底、约定双份人肉对齐、命名无 convention 自然漂移 |
|
||||
| 复用仓内 JSON-RPC 2.0（dsh-sdk-jsonrpc-server） | 数字错误码退化成单码兜底、约定双份人肉对齐、命名无 convention 自然漂移 |
|
||||
| 三信封模型（Request/Response/Frame 各一信封，签名不感知方向） | rpcId 是逻辑层关联，帧与应答的方向语义靠通道推断在换载体时即失效 |
|
||||
| 具名 Request/Response 类型对为真源（map 登记类型对） | 平铺具名类型是同一事实的第二个名字；签名 infer 反推让加方法只改一处 |
|
||||
| REST 风格路径 | 消费方是自家 client，无第三方 REST 体验诉求；RPC 直映方法表更机械 |
|
||||
|
||||
@@ -2,5 +2,5 @@
|
||||
# side as of the last confirmed-consistent state. Both languages carry equal authority;
|
||||
# after editing either side, bring the other along and re-record with:
|
||||
# pnpm run verify-translation-pairing --write .agents/notes/implemented/architecture/2026-07-19-package-invariant-runtime-contracts.md
|
||||
2026-07-19-package-invariant-runtime-contracts.md: 6e6516c2fa629aa736c178be4acbc8b42fd9a0b9
|
||||
2026-07-19-package-invariant-runtime-contracts.zh.md: 6e6da337d811a25f073b424af8dc3bbc483520a6
|
||||
2026-07-19-package-invariant-runtime-contracts.md: b5799a37a61244193b46db6ea4ae15f306d144b2
|
||||
2026-07-19-package-invariant-runtime-contracts.zh.md: e6035bbabba7188017746c57c5b6a48761710658
|
||||
|
||||
@@ -39,17 +39,17 @@ The current 103-package workspace has 21 executable companions and 82 justified
|
||||
| `dsh-llm-retry` | Durable retry records identify the open turn's latest closed step, remain unique per step, increase monotonically, and stay within retry and non-negative timer bounds. |
|
||||
| `dsh-tools` | Monotonic pre/execute/post stages and immutable final execution/result snapshots. |
|
||||
| `dsh-system-prompt` | Authoritative assembly section, tool, and variable data constraints. |
|
||||
| `dsh-compact` | Compaction start/summary/end pairing, range endpoints, token counts, and successful-summary presence. |
|
||||
| `dsh-compaction` | Compaction start/summary/end pairing, range endpoints, token counts, and successful-summary presence. |
|
||||
| `dsh-hook-protocol` | Hook invocation/result correlation, dialect, identity, and duration constraints. |
|
||||
| `dsh-sandbox-policy` | Durable `sandbox/mode` events use the closed sandbox-mode vocabulary. |
|
||||
| `dsh-fs` | Filesystem decision/observation events carry usable target and version identities. |
|
||||
| `dsh-goal` | Durable goal snapshots preserve source attribution, rendered content, revisions, lifecycle and timestamp relationships, and sequential admitted rounds. |
|
||||
| `dsh-goal-session` | Goal-sourced continuation messages match the prompt reconstructed from the preceding durable goal state. |
|
||||
| `dsh-goal-round-driver` | Goal-sourced continuation messages match the prompt reconstructed from the preceding durable goal state. |
|
||||
| `dsh-subagent` | Provider add/remove and child start/end events preserve identity and pairing. |
|
||||
| `dsh-permission` | Durable permission decisions name a preset in the active permission table. |
|
||||
| `dsh-permission-presets` | Durable permission decisions name a preset in the active permission table. |
|
||||
| `dsh-user-approval` | Approval asked/decided records pair by call and use valid outcomes and policies. |
|
||||
| `dsh-workflow` | Workflow and child-agent start/end events preserve run metadata, identity, outcome, count, and error relations. |
|
||||
| `dsh-tasks` | Current and terminal task snapshots preserve id/kind, owner, status, and timestamp relationships. |
|
||||
| `dsh-jobs` | Current and terminal task snapshots preserve id/kind, owner, status, and timestamp relationships. |
|
||||
| `dsh-tool-todo` | Durable whole-list snapshots use unique trimmed items and closed statuses. |
|
||||
| `dsh-time-context` | Plugin-attributed clock readings agree with the session's open turn, next pre-step position, and elapsed baseline; rendered time parses and does not postdate its event. |
|
||||
|
||||
@@ -59,7 +59,7 @@ Session-backed companions validate existing durable events when they load, using
|
||||
|
||||
`verify-package-invariants` discovers every workspace package and enforces companion source, exact-name registration, named-only Loader shape, `./invariant` exports, publication files, dependencies, TypeScript references, and bundle entries. Its AST rule rejects generated markers, default exports, and unexplained empty installers. A non-empty installer must accept and use the failure reporter, and registration must pass that checked local `install` function. The gate deliberately does not infer semantic quality from method names or helper calls.
|
||||
|
||||
Vitest mounts `InvariantService` with `{ enabled: true }` for every package test topology and loads the owning companion. The invariant subpath path mapping resolves source companions instead of stale built output. Focused suites cover every executable companion's valid and invalid observations, and the exhaustive topology runs every source companion through the real Loader namespace normalization. After the structural gate validates each publication map, an artifact gate stages its manifest-declared `lib/` files, imports the compiled `./invariant` self-reference under plain Node, and repeats that Loader-shape check, so a companion that imports an undeclared runtime chunk fails before release. Tests that synthesize event streams must produce a valid surrounding lifecycle unless the test is intentionally asserting a violation.
|
||||
Vitest mounts `InvariantRegistry` with `{ enabled: true }` for every package test topology and loads the owning companion. The invariant subpath path mapping resolves source companions instead of stale built output. Focused suites cover every executable companion's valid and invalid observations, and the exhaustive topology runs every source companion through the real Loader namespace normalization. After the structural gate validates each publication map, an artifact gate stages its manifest-declared `lib/` files, imports the compiled `./invariant` self-reference under plain Node, and repeats that Loader-shape check, so a companion that imports an undeclared runtime chunk fails before release. Tests that synthesize event streams must produce a valid surrounding lifecycle unless the test is intentionally asserting a violation.
|
||||
|
||||
## Alternatives considered
|
||||
|
||||
|
||||
@@ -39,17 +39,17 @@ Status: implemented
|
||||
| `dsh-llm-retry` | 持久化重试记录指向当前打开轮次中最近关闭的步骤；每个步骤的记录保持唯一，重试次数单调递增，并且重试次数和非负的定时器延迟均保持在边界内。 |
|
||||
| `dsh-tools` | pre/execute/post 阶段单调推进，以及最终 execution/result 快照不可变。 |
|
||||
| `dsh-system-prompt` | 权威 assembly 中 section、工具和 variable 的数据约束。 |
|
||||
| `dsh-compact` | 压缩（compaction）start/summary/end 配对、范围端点、token 数量和成功时必须存在 summary。 |
|
||||
| `dsh-compaction` | 压缩（compaction）start/summary/end 配对、范围端点、token 数量和成功时必须存在 summary。 |
|
||||
| `dsh-hook-protocol` | 钩子 invocation/result 的关联、dialect、身份和 duration 约束。 |
|
||||
| `dsh-sandbox-policy` | 持久化 `sandbox/mode` 事件必须使用封闭的 sandbox-mode 词表。 |
|
||||
| `dsh-fs` | 文件系统决策/观测事件必须携带可用的 target 和 version 身份。 |
|
||||
| `dsh-goal` | 持久化目标快照保持来源归属、渲染内容、修订号、生命周期和时间戳关系，并保证已准入的 Round 连续编号。 |
|
||||
| `dsh-goal-session` | 目标来源的继续执行消息必须匹配根据此前持久化目标状态重建的提示词。 |
|
||||
| `dsh-goal-round-driver` | 目标来源的继续执行消息必须匹配根据此前持久化目标状态重建的提示词。 |
|
||||
| `dsh-subagent` | 提供方 add/remove 和 child start/end 事件必须保持身份与配对。 |
|
||||
| `dsh-permission` | 持久化 permission 决策必须引用当前 permission 表中的 preset。 |
|
||||
| `dsh-permission-presets` | 持久化 permission 决策必须引用当前 permission 表中的 preset。 |
|
||||
| `dsh-user-approval` | approval asked/decided 记录按 call 配对，并使用有效 outcome 和 policy。 |
|
||||
| `dsh-workflow` | 工作流和 child-agent start/end 事件保持 run metadata、身份、outcome、数量和 error 关系。 |
|
||||
| `dsh-tasks` | 当前与终态 task 快照保持 id/kind、owner、status 和 timestamp 关系。 |
|
||||
| `dsh-jobs` | 当前与终态 task 快照保持 id/kind、owner、status 和 timestamp 关系。 |
|
||||
| `dsh-tool-todo` | 持久化全量快照使用唯一且已 trim 的条目和封闭 status。 |
|
||||
| `dsh-time-context` | 标注插件来源的时钟 reading 必须匹配会话当前打开的轮次、下一个步骤开始前的位置和 elapsed baseline；渲染时间必须可解析，且不得晚于对应事件。 |
|
||||
|
||||
@@ -59,7 +59,7 @@ Status: implemented
|
||||
|
||||
`verify-package-invariants` 发现每个 workspace 包，并强制 companion 源文件、完整名称注册、仅含具名 export 的 Loader 形状、`./invariant` export、发布文件、依赖、TypeScript reference 和 bundle entry 完整。其 AST 规则拒绝生成标记、默认导出和没有解释的空安装器。非空安装器必须接收并使用失败报告器，注册时还必须传入该经检查的本地 `install` 函数。门禁不会通过方法名或 helper 调用推断语义质量。
|
||||
|
||||
Vitest 为每个包测试拓扑使用 `{ enabled: true }` 挂载 `InvariantService`，并加载所有者 companion。不变量 subpath 的 path mapping 会解析源 companion，而不是陈旧的构建输出。聚焦 suite 覆盖每个可执行 companion 的有效和无效观测；穷举拓扑通过真实 Loader 命名空间归一化运行每个源 companion。结构门禁验证每个包的发布映射后，产物门禁会暂存其 manifest（元数据清单）声明的 `lib/` 文件，在 plain Node 下导入已编译的 `./invariant` 自引用，并重复执行该 Loader 形状检查；这样，若 companion 导入未声明的运行时分片，门禁就会在发布前失败。合成事件流的测试必须构造有效的外围生命周期，除非测试本身就是在断言违规。
|
||||
Vitest 为每个包测试拓扑使用 `{ enabled: true }` 挂载 `InvariantRegistry`，并加载所有者 companion。不变量 subpath 的 path mapping 会解析源 companion，而不是陈旧的构建输出。聚焦 suite 覆盖每个可执行 companion 的有效和无效观测；穷举拓扑通过真实 Loader 命名空间归一化运行每个源 companion。结构门禁验证每个包的发布映射后，产物门禁会暂存其 manifest（元数据清单）声明的 `lib/` 文件，在 plain Node 下导入已编译的 `./invariant` 自引用，并重复执行该 Loader 形状检查；这样，若 companion 导入未声明的运行时分片，门禁就会在发布前失败。合成事件流的测试必须构造有效的外围生命周期，除非测试本身就是在断言违规。
|
||||
|
||||
## 考虑过的替代方案
|
||||
|
||||
|
||||
@@ -2,5 +2,5 @@
|
||||
# side as of the last confirmed-consistent state. Both languages carry equal authority;
|
||||
# after editing either side, bring the other along and re-record with:
|
||||
# pnpm run verify-translation-pairing --write .agents/notes/implemented/architecture/2026-07-20-canonical-tool-output-contract.md
|
||||
2026-07-20-canonical-tool-output-contract.md: fc4a8d52f2ad6ff532fafdd7a987a095c8e3bcef
|
||||
2026-07-20-canonical-tool-output-contract.zh.md: 034e2481cf532268dc037db867e3a332e7a8d5c9
|
||||
2026-07-20-canonical-tool-output-contract.md: 0b3bd788fd1ab63aa6929827ef82e5ba732e5324
|
||||
2026-07-20-canonical-tool-output-contract.zh.md: bbcd519de8d02df14be931b44f8dc44fc06b8f6a
|
||||
|
||||
@@ -47,10 +47,10 @@ The first-party tools preserve their existing Native text while returning domain
|
||||
| `grep` | `{ matches: [{ path, lineNumber, line }] }` |
|
||||
| `web_search` / `web_fetch` | The normalized `WebSearchResult` / `WebFetchResult` |
|
||||
| `lsp` | `{ kind: "locations", locations, resolvedWorkspaceUri }` or `{ kind: "hover", hover }` |
|
||||
| `bash` | `{ kind: "background", taskId }` or `{ kind: "foreground" } & BashRunResult` |
|
||||
| `terminal_open` / `terminal_list` / `terminal_send` / `terminal_read` / `terminal_signal` / `terminal_close` | Public session snapshots, bounded read/send DTOs, signal/close outcomes, or a background task handle |
|
||||
| `task_output` / `task_list` / `task_kill` | Public task snapshots without owner or notification bookkeeping |
|
||||
| `subagent` | Background task handle or `{ kind: "foreground", runId, output: JsonValue[] }` |
|
||||
| `bash` | `{ kind: "background", jobId }` or `{ kind: "foreground" } & ShellRunResult` |
|
||||
| `terminal_open` / `terminal_list` / `terminal_send` / `terminal_read` / `terminal_signal` / `terminal_close` | Public session snapshots, bounded read/send DTOs, signal/close outcomes, or a background job handle |
|
||||
| `job_output` / `job_list` / `job_kill` | Public task snapshots without owner or notification bookkeeping |
|
||||
| `subagent` | Background job handle or `{ kind: "foreground", runId, output: JsonValue[] }` |
|
||||
| `workflow` / `ralph` | `{ runId, agentsStarted, result: JsonValue }` |
|
||||
| `skill` | `{ name, provider, resourceBase?, content }` |
|
||||
| `todo_write` | `{ todos, counts }` |
|
||||
@@ -66,7 +66,7 @@ MCP bridges preserve protocol blocks through `McpResult<{...}> = { content: Json
|
||||
|
||||
## Alternatives considered
|
||||
|
||||
- **Return rendered text to Code Mode:** rejected because callers would continue scraping prose for task ids, mount ids, paths, and structured provider results.
|
||||
- **Return rendered text to Code Mode:** rejected because callers would continue scraping prose for job ids, mount ids, paths, and structured provider results.
|
||||
- **Persist canonical values on `tool/result`:** rejected because nested execution values are not model history, need not survive replay, and would create a session-format and storage commitment unrelated to Native reconstruction.
|
||||
- **Let tools return both value and content:** rejected because two author-owned results can disagree and policy cannot state which one is authoritative. The renderer makes presentation a deterministic projection of the validated value.
|
||||
- **Treat content replacement as value redaction:** rejected because presentation and programmatic access are different consumers; hiding only the former would create a false security boundary.
|
||||
|
||||
@@ -47,9 +47,9 @@ type ToolExecutionResult =
|
||||
| `grep` | `{ matches: [{ path, lineNumber, line }] }` |
|
||||
| `web_search` ／ `web_fetch` | 归一化后的 `WebSearchResult` ／ `WebFetchResult` |
|
||||
| `lsp` | `{ kind: "locations", locations, resolvedWorkspaceUri }` 或 `{ kind: "hover", hover }` |
|
||||
| `bash` | `{ kind: "background", taskId }` 或 `{ kind: "foreground" } & BashRunResult` |
|
||||
| `bash` | `{ kind: "background", jobId }` 或 `{ kind: "foreground" } & ShellRunResult` |
|
||||
| `terminal_open` ／ `terminal_list` ／ `terminal_send` ／ `terminal_read` ／ `terminal_signal` ／ `terminal_close` | 公开会话快照、有界的读取／发送 DTO、信号／关闭操作结果，或后台任务句柄 |
|
||||
| `task_output` ／ `task_list` ／ `task_kill` | 不含所有者或通知管理信息的公开任务快照 |
|
||||
| `job_output` ／ `job_list` ／ `job_kill` | 不含所有者或通知管理信息的公开任务快照 |
|
||||
| `subagent` | 后台任务句柄或 `{ kind: "foreground", runId, output: JsonValue[] }` |
|
||||
| `workflow` ／ `ralph` | `{ runId, agentsStarted, result: JsonValue }` |
|
||||
| `skill` | `{ name, provider, resourceBase?, content }` |
|
||||
@@ -66,7 +66,7 @@ MCP 桥接层通过 `McpResult<{...}> = { content: JsonValue[]; structuredConten
|
||||
|
||||
## 备选方案
|
||||
|
||||
- **向 Code Mode 返回渲染后的文本：**不予采纳。调用方仍需从自然语言中提取 task id、挂载 id、路径和结构化提供方结果。
|
||||
- **向 Code Mode 返回渲染后的文本：**不予采纳。调用方仍需从自然语言中提取 job id、挂载 id、路径和结构化提供方结果。
|
||||
- **在 `tool/result` 上持久化规范值：**不予采纳。嵌套执行值不属于模型历史记录，无需在回放后继续存在；持久化还会引入与 Native 重建无关的会话格式和存储承诺。
|
||||
- **允许工具同时返回值和内容：**不予采纳。由作者分别维护的两份结果可能互相矛盾，策略也无法说明哪一份才是权威结果。渲染器会根据已校验值确定性地产生展示。
|
||||
- **将内容替换视为值脱敏：**不予采纳。展示内容和程序化访问面向不同消费方；只隐藏前者会制造虚假的安全边界。
|
||||
|
||||
@@ -2,5 +2,5 @@
|
||||
# side as of the last confirmed-consistent state. Both languages carry equal authority;
|
||||
# after editing either side, bring the other along and re-record with:
|
||||
# pnpm run verify-translation-pairing --write .agents/notes/implemented/architecture/2026-07-20-routed-model-context-and-compaction-policy.md
|
||||
2026-07-20-routed-model-context-and-compaction-policy.md: 630c259e5bbf6edc46cc837aedb627e673dc71f2
|
||||
2026-07-20-routed-model-context-and-compaction-policy.zh.md: c874a0fed4f07066e2fc5eec107eb1701918f7a3
|
||||
2026-07-20-routed-model-context-and-compaction-policy.md: ca898acefa4dde9190e0d17249ec07cc875d454f
|
||||
2026-07-20-routed-model-context-and-compaction-policy.zh.md: a55be759da19cb432c8dda360eea7cdd641a47b6
|
||||
|
||||
@@ -14,19 +14,19 @@ Neither obvious configuration owner is sufficient. Compact-basic is optional and
|
||||
|
||||
### Adapters own exact-route capacity
|
||||
|
||||
`LlmAdapter.resolveModel(provider, model, signal?)` returns aggregate metadata for one exact route, with optional `LlmModelContext` under its `context` field. `LlmService.resolveModelInfo()` selects the registered route owner, validates a positive integer `contextWindow`, and returns detached metadata. The query is independent of `listModels()`: an unlisted dynamic model may have capacity metadata, and an absent `context` means only that the adapter cannot describe capacity.
|
||||
`LlmAdapter.resolveModel(provider, model, signal?)` returns aggregate metadata for one exact route, with optional `LlmModelContext` under its `context` field. `LlmRuntime.resolveModelInfo()` selects the registered route owner, validates a positive integer `contextWindow`, and returns detached metadata. The query is independent of `listModels()`: an unlisted dynamic model may have capacity metadata, and an absent `context` means only that the adapter cannot describe capacity.
|
||||
|
||||
The hand-rolled DeepSeek adapter accepts optional `contextWindow` on each configured model plus an adapter-wide `defaultContextWindow`. Exact model capacity wins; an entry without capacity and an unlisted pass-through id inherit the adapter default, or omit `context` when it is absent. The two built-in model entries each publish an exact 256,000-token capacity. The pi-ai adapter resolves capacity from the same catalog descriptor that authoritatively resolves the request model.
|
||||
|
||||
### Token measurement remains model-agnostic
|
||||
|
||||
`dsh-token-meter` has no configuration and no model profiles. It owns one fixed replay fold and returns absolute estimated token pressure plus positional surface prices. Removing global capacity keeps measurement reusable when compact-basic is absent and prevents replay accounting from becoming another model registry.
|
||||
`dsh-token-meter` has no configuration and no model profiles. It owns one fixed replay fold and returns absolute estimated token pressure plus positional surface prices. Removing global capacity keeps measurement reusable when compaction-basic is absent and prevents replay accounting from becoming another model registry.
|
||||
|
||||
### Compact-basic resolves a target spec
|
||||
|
||||
Compact-basic owns consumer policy. Top-level fields define defaults; `modelPolicies` contains partial overrides keyed by the exact `{ provider, model }` pair. Duplicate targets and unknown or invalid fields fail plugin load. `thresholdRatio` defaults to `0.8`, and retention defaults to `retainRatio: 0.16`; callers may use an absolute `retainTokens` instead, but the two retention forms are mutually exclusive. After inheritance, a ratio retention that is not below its threshold ratio also fails plugin load because no model capacity can make that policy valid.
|
||||
|
||||
For proactive pressure, compact-basic reads the latest durable request route, resolves its adapter capacity and exact-target policy, and scales ratios into a `ResolvedCompactSpec`. It performs this resolution on every check, so a provider or model switch in one session changes capacity and policy immediately. An absolute retained budget that is not below the scaled threshold fails when the target capacity first makes that comparison possible.
|
||||
For proactive pressure, compaction-basic reads the latest durable request route, resolves its adapter capacity and exact-target policy, and scales ratios into a `ResolvedCompactSpec`. It performs this resolution on every check, so a provider or model switch in one session changes capacity and policy immediately. An absolute retained budget that is not below the scaled threshold fails when the target capacity first makes that comparison possible.
|
||||
|
||||
The same exact-target override can select summarization provider/model, summarization output cap, convergence retries, and overflow retry cap. These are compaction concerns and never enter an LLM provider.
|
||||
|
||||
@@ -40,7 +40,7 @@ Service tests cover detached context metadata, invalid adapter output, catalog i
|
||||
|
||||
## Alternatives considered
|
||||
|
||||
- **Put capacity and all policies in compact-basic** — rejected because compact-basic would duplicate adapter model knowledge, dynamic unlisted models would require parallel registration, and capacity would disappear when compaction is not installed.
|
||||
- **Put capacity and all policies in compaction-basic** — rejected because compaction-basic would duplicate adapter model knowledge, dynamic unlisted models would require parallel registration, and capacity would disappear when compaction is not installed.
|
||||
- **Put compaction policy in each LLM adapter** — rejected because adapters must remain independent of optional consumers, while summarization and retry policy are not provider facts.
|
||||
- **Make `listModels()` authoritative** — rejected because discovery is advisory and some adapters intentionally accept dynamic ids. Correctness metadata must not turn selector membership into a routing whitelist.
|
||||
- **Add per-model folds to token-meter** — rejected because the replay algorithm is shared; only the capacity and consumer policy change. Multiple folds would duplicate state without improving estimation.
|
||||
@@ -49,8 +49,8 @@ Service tests cover detached context metadata, invalid adapter output, catalog i
|
||||
## Consequences
|
||||
|
||||
- Capacity has one authoritative owner at the provider contract, while compaction policy stays in the optional consuming plugin.
|
||||
- The same compact-basic instance safely handles different windows, provider switches, and identical model ids under different providers without consulting discovery metadata.
|
||||
- LLM-only and meter-only compositions remain valid; loading compact-basic adds no reverse dependency from adapters.
|
||||
- The same compaction-basic instance safely handles different windows, provider switches, and identical model ids under different providers without consulting discovery metadata.
|
||||
- LLM-only and meter-only compositions remain valid; loading compaction-basic adds no reverse dependency from adapters.
|
||||
- DeepSeek deployments may set exact per-model capacities, or use `defaultContextWindow` for entries without capacity and unlisted pass-through ids.
|
||||
- Ratio defaults scale naturally across models, while exact-target absolute retention remains available for deployment-specific behavior.
|
||||
|
||||
|
||||
@@ -14,19 +14,19 @@ Status: implemented
|
||||
|
||||
### 适配器拥有精确路由容量
|
||||
|
||||
`LlmAdapter.resolveModel(provider, model, signal?)` 返回一条精确路由的聚合元数据，其中可选的 `LlmModelContext` 位于 `context` 字段下。`LlmService.resolveModelInfo()` 选择已注册的路由所属方，验证 `contextWindow` 为正整数，并返回与适配器内部状态分离的元数据。该查询独立于 `listModels()`：不在目录中的动态模型也可以拥有容量元数据，而缺少 `context` 只表示适配器无法描述容量。
|
||||
`LlmAdapter.resolveModel(provider, model, signal?)` 返回一条精确路由的聚合元数据，其中可选的 `LlmModelContext` 位于 `context` 字段下。`LlmRuntime.resolveModelInfo()` 选择已注册的路由所属方，验证 `contextWindow` 为正整数，并返回与适配器内部状态分离的元数据。该查询独立于 `listModels()`：不在目录中的动态模型也可以拥有容量元数据，而缺少 `context` 只表示适配器无法描述容量。
|
||||
|
||||
手写 DeepSeek 适配器允许每个已配置模型提供可选 `contextWindow`，并支持适配器级 `defaultContextWindow`。精确模型容量优先；未提供容量的模型项与未列出的透传 id 会继承适配器默认值，若默认值也不存在则省略 `context`。两个内置模型项都公开精确的 256,000 token 容量。pi-ai 适配器从同一个目录描述符解析容量，该描述符也用于权威解析请求模型。
|
||||
|
||||
### Token 计量保持模型无关
|
||||
|
||||
`dsh-token-meter` 没有配置，也没有模型 profile。它拥有一个固定回放折叠，并返回绝对估算 token 压力，以及按位置排列的表层节点 token 估值。移除全局容量后，未加载 compact-basic 时仍可复用计量，同时避免让回放核算变成另一套模型注册表。
|
||||
`dsh-token-meter` 没有配置，也没有模型 profile。它拥有一个固定回放折叠，并返回绝对估算 token 压力，以及按位置排列的表层节点 token 估值。移除全局容量后，未加载 compaction-basic 时仍可复用计量，同时避免让回放核算变成另一套模型注册表。
|
||||
|
||||
### Compact-basic 解析目标规格
|
||||
|
||||
Compact-basic 拥有消费方策略。顶层字段定义默认值；`modelPolicies` 包含以精确 `{ provider, model }` 组合为键的部分覆盖。重复目标、未知字段或无效字段都会让插件加载失败。`thresholdRatio` 默认为 `0.8`，保留策略默认为 `retainRatio: 0.16`；调用方也可以改用绝对 `retainTokens`，但两种保留形式互斥。完成继承后，如果保留比例不小于阈值比例，插件也会加载失败，因为任何模型容量都无法让该策略有效。
|
||||
|
||||
对于主动压力检查，compact-basic 读取最新持久请求路由，解析其适配器容量与精确目标策略，再把比例缩放为 `ResolvedCompactSpec`。每次检查都会重新解析，因此同一会话切换提供方或模型后，容量与策略会立即变化。若绝对保留预算不小于缩放后的阈值，系统会在目标容量首次允许比较两者时失败。
|
||||
对于主动压力检查，compaction-basic 读取最新持久请求路由，解析其适配器容量与精确目标策略，再把比例缩放为 `ResolvedCompactSpec`。每次检查都会重新解析，因此同一会话切换提供方或模型后，容量与策略会立即变化。若绝对保留预算不小于缩放后的阈值，系统会在目标容量首次允许比较两者时失败。
|
||||
|
||||
同一精确目标覆盖还可以选择摘要提供方/模型、摘要输出上限、收敛重试次数与溢出重试上限。这些都属于压缩问题，不会进入任何 LLM 提供方。
|
||||
|
||||
@@ -40,7 +40,7 @@ Compact-basic 拥有消费方策略。顶层字段定义默认值；`modelPolici
|
||||
|
||||
## 考虑过的替代方案
|
||||
|
||||
- **把容量与所有策略都放进 compact-basic**——不予采纳，因为 compact-basic 会复制适配器的模型知识，未列出的动态模型需要并行注册，而且未安装压缩时容量也会消失。
|
||||
- **把容量与所有策略都放进 compaction-basic**——不予采纳，因为 compaction-basic 会复制适配器的模型知识，未列出的动态模型需要并行注册，而且未安装压缩时容量也会消失。
|
||||
- **把压缩策略放进各个 LLM 适配器**——不予采纳，因为适配器必须独立于可选消费方，而摘要与重试策略也不是提供方事实。
|
||||
- **让 `listModels()` 成为权威来源**——不予采纳，因为发现能力只是建议信息，一些适配器有意接受动态 id。正确性元数据不能把选择器成员关系变成路由白名单。
|
||||
- **给 token-meter 增加逐模型折叠**——不予采纳，因为回放算法可以共享，变化的只有容量与消费方策略。多个折叠会重复状态，却不会改善估算。
|
||||
@@ -49,8 +49,8 @@ Compact-basic 拥有消费方策略。顶层字段定义默认值；`modelPolici
|
||||
## 后果
|
||||
|
||||
- 容量在提供方约定上拥有唯一权威归属方，而压缩策略留在可选消费插件中。
|
||||
- 同一个 compact-basic 实例无需查询发现元数据，就能安全处理不同窗口、提供方切换，以及不同提供方下的相同模型 id。
|
||||
- 仅 LLM 与仅 meter 的组合仍然有效；加载 compact-basic 不会让适配器产生反向依赖。
|
||||
- 同一个 compaction-basic 实例无需查询发现元数据，就能安全处理不同窗口、提供方切换，以及不同提供方下的相同模型 id。
|
||||
- 仅 LLM 与仅 meter 的组合仍然有效；加载 compaction-basic 不会让适配器产生反向依赖。
|
||||
- DeepSeek 部署可以设置精确的逐模型容量，也可以让未提供容量的模型项与未列出的透传 id 使用 `defaultContextWindow`。
|
||||
- 比例默认值会随模型自然缩放，同时仍可按精确目标使用绝对保留值，以满足部署专用行为。
|
||||
|
||||
|
||||
@@ -2,5 +2,5 @@
|
||||
# side as of the last confirmed-consistent state. Both languages carry equal authority;
|
||||
# after editing either side, bring the other along and re-record with:
|
||||
# pnpm run verify-translation-pairing --write .agents/notes/implemented/architecture/2026-07-22-slot-type-chain-implementation.md
|
||||
2026-07-22-slot-type-chain-implementation.md: 41a5c4592b22cc66d2f717794534305972c89eb3
|
||||
2026-07-22-slot-type-chain-implementation.zh.md: f1be2d1f62b1bc9df6b14d6824c489d44cc9eca5
|
||||
2026-07-22-slot-type-chain-implementation.md: 48c3319cdab26b2e053cb08ed1dafa2b96a980d7
|
||||
2026-07-22-slot-type-chain-implementation.zh.md: 496677ecdef7d5b691edfc0bf43c9ff9d3d80449
|
||||
|
||||
@@ -16,7 +16,7 @@ One sentence: **the shell renders only `'root'`; a plugin composes UI through a
|
||||
|
||||
### 'root' is the only a-priori slot
|
||||
|
||||
`SlotsService` (client runtime) declares `'root'` at construction — single/root, `owner: {}` — and its `SlotMap` merge lives in the runtime package. The shell's entire assembly is `ctx.slots.renderSlot('root', {})`: the only ctx-level render entry; any other key, a missing renderer, or an unregistered root fails loud (no fallback).
|
||||
`SlotRegistry` (client runtime) declares `'root'` at construction — single/root, `owner: {}` — and its `SlotMap` merge lives in the runtime package. The shell's entire assembly is `ctx.slots.renderSlot('root', {})`: the only ctx-level render entry; any other key, a missing renderer, or an unregistered root fails loud (no fallback).
|
||||
|
||||
### register is the single API; children = declaration + authorization + runtime spec
|
||||
|
||||
|
||||
Some files were not shown because too many files have changed in this diff Show More
Reference in New Issue
Block a user