Commit Graph

82 Commits

Author SHA1 Message Date
Turtle
0c708cb10d refactor: replace overloaded surface terminology 2026-08-11 15:23:05 +08:00
Hypatia May
85dd22fd4f feat(subagent): deliver continuable child settlement to parents
A continuable child that stopped without reporting — an error, a token
ceiling, cancellation, teardown — left its parent nothing to act on.
The continuation manager now delivers an unconditional settlement
notice to the durable direct parent before releasing ownership, folding
consumed work (foldConsumedWork supersedes findLastMessageTurnEnd) so a
claimed-but-unrun prompt reads as aborted rather than completed, waking
an idle parent, steering a busy one, and never waking a closing tree.
2026-08-11 12:31:49 +08:00
Hypatia May
f27bdb2be3 Merge remote-tracking branch 'origin/master' into fix/subagent-empty-terminal-message-output
# Conflicts:
#	docs/event-producer-consumer.i18n.yaml
#	docs/event-producer-consumer.md
#	docs/event-producer-consumer.zh.md
#	docs/subsystems/subagent.i18n.yaml
#	docs/subsystems/subagent.md
#	docs/subsystems/subagent.zh.md
2026-08-10 22:59:23 +08:00
Hypatia May
29c00d59a5 Merge remote-tracking branch 'origin/master' into fix/continuable-subagent-policy-inheritance
# Conflicts:
#	packages/subagent/subagent-inprocess/package.json
#	packages/subagent/subagent/package.json
#	pnpm-lock.yaml
2026-08-10 22:24:56 +08:00
Hypatia May
eebece5e62 Merge remote-tracking branch 'origin/master' into fix/subagent-empty-terminal-message-output 2026-08-10 22:19:44 +08:00
imccyu
ec601ca13d build(vendor): rescope the vendored Cordis packages into @deepseek-ai
Machine-produced by `pnpm run rescope-vendor --apply` plus the regeneration it
prints: `pnpm install` for the lockfile, `pnpm run gen-third-party-notices`,
`verify-translation-pairing --write` for the touched bilingual pairs,
`gen-doc-graphs`, and one typert snapshot whose ids embed character offsets.
`pnpm run rescope-vendor --check` verifies the result.

Renames nine vendored packages (cordis, cosmokit, schemastery and the six
@cordisjs plugins) and every reference that resolves them: manifest names and
dependency keys, module specifiers including declare-module merges, cordis.yml
plugin names, tsconfig paths, every Markdown fence, and `docs/` prose.
Directory names, upstream versions, and dependency ranges are unchanged, so
vendor/README.md still reads as an upstream snapshot; its manifest table gains
an upstream-name column so THIRD_PARTY_NOTICES keeps MIT attribution pointed
at each fork's origin.

The tutorial tier follows the rename end to end: its yaml fences named plugins
the Loader can no longer resolve, its `ts ignore-check` fences disagreed with
the compiled fences beside them, and its prose quoted both. The contracts that
told readers to keep upstream names — the root convention and the vendoring
cookbook's tree comment and manifest invariant — now say to rescope instead.

Two rules read `@deepseek-ai/` as "another workspace plugin": the client bundle
purity gate now names the vendored libraries a browser bundle inlines, and the
files where a bare `cordis` is an agent-preset id keep that product data.
2026-08-10 22:04:13 +08:00
Hypatia May
ea47c32805 review: one fold implementation, uniform end-edge absence, partial text in tool errors, snapshot scenario
Address ds-review-bot on #2127:

- assistant-output: the rule has ONE implementation, the incremental
  AssistantOutputFold (push/pushText/collect); finalAssistantOutput folds a
  complete suffix, the SDK backend folds notification events, and the ACP
  backend folds raw chunk text into the same streamed fallback.
- subagent/end.lastAssistantMessage: 'no output' is encoded once — absent,
  never [], on both lifecycle shapes (observeRun now omits empty output).
- tool-subagent: a non-completed foreground result stays isError but appends
  the child's preserved partial text after the stop-reason headline.
- Authored keyless snapshot scenario subagent-max-tokens-partial pins the
  assembled transcript: the child's committed log carries the usage-only
  empty message and the parent's tool result carries the partial answer.
- Rule-boundary sentence (message wins over later streamed text) and the
  consumer half recorded in the Agent Note; comments trimmed to pointers.
2026-08-10 16:09:04 +08:00
Hypatia May
64e0fbfd6d fix(subagent): inherit parent policy overrides in continuable children
A continuable background child (the default backgroundMode for both
delegation tools) never received the parent session's explicit
sandbox/approval overrides: materialization applied only child
composition, so a danger-full-access parent produced workspace-write
children whose every out-of-workspace operation raised an approval
prompt.

Move the one-shot driver's capture/append pair into the shared
child-agent module (captureDelegatedPolicyOverrides /
appendDelegatedPolicyOverrides) and call it from both paths:
startContinuable captures before its first await, only fresh
materialization appends the source-tagged events (after any fork seed),
and a cold resume replays the persisted delegation events instead of
re-capturing the parent.

Adds the continuable inheritance unit suite, the ACP snapshot scenario
subagent-continuable-inheritance (fails without the fix), the
continuable policy-inheritance Agent Note, and the seam-level README
contract, with bilingual counterparts.

Fixes #1692
2026-08-10 12:16:19 +08:00
Hypatia May
1db1cda464 fix(subagent): keep output past an empty terminal message with one selection rule
A max-tokens step that assembled only tool-call blocks appends an
EMPTY-content assistant/message (the usage host). Three consumers each
hand-rolled output selection and all let it erase the child's real
answer: the in-process readResult and the Activation subagent/end
capture took the last message unfiltered, and the SDK backend let any
message beat its streamed-text fallback; the in-process driver also had
no streamed-text fallback for cancelled turns.

dsh-subagent now owns the canonical rule in src/assistant-output.ts
(last non-empty assistant message, else the accumulated text-delta
stream) and all three consumers apply it. Regression tests in all three
packages fail under the previous selections.

Closes #1514
2026-08-10 11:16:29 +08:00
Turtle
dda02250f5 docs: reserve seam for complete capabilities 2026-08-09 17:26:57 +08:00
Turtle
9704749b01 docs: replace vague provenance prose with recorded facts 2026-08-09 15:35:02 +08:00
Hypatia May
d769d3cbb7 feat(subagent): let ancestors interrupt descendants
interrupt_agent(agent_id) passes the calling agent as the ancestor
authority for ctx.subagents.interrupt(); the service verifies live
registry identity and recorded lineage, so a direct child or deeper
descendant stops with the same generic parameter while send_message keeps
its exact-direct-parent authority.

Discovery: list_agents gains an optional scope. descendants walks the new
SubagentService.listDescendants() — one lineage trace flattened in stable
pre-order across ordinary and one-shot intermediates, each entry carrying
its verified parentId and depth — and every status now comes from the
live Agent registry (running/idle/complete).

Refs #1535
2026-08-08 20:02:37 +08:00
Hypatia May
336c84baf0 docs(subagent): clarify interrupt edge cases 2026-08-08 20:02:36 +08:00
Hypatia May
66a21a38b1 feat(subagent): add current-turn interrupt RPC
ctx.subagents.interrupt() stops one live continuable child's current turn
via Agent.cancel(cause, { keepInbox: true }) under either a human durable
parent address or an exact live ancestor Agent. Fire-and-return: admission
is synchronous, quiescence is not awaited. Pending inbox work, the
Activation, and published descendants are preserved; only a later waking
send resumes the parked FIFO queue. Absent, one-shot, and disposing
targets are accepted no-ops.

The new Host RPC subagent.interrupt calls only that primitive with user
authority — no catalog, history, persistence, or parent-registry lookup —
so a live child stays stoppable while its parent Agent is offline.

Refs #1535
2026-08-08 20:02:36 +08:00
imccyu
c98a754ccb fix: address review round four
- the cached-identity rung gains a finality gate: the identity value
  carries its descriptor seq and a cached row is served only when that
  seq lands in the child's own suffix, so a fork seed's replayed
  ancestor identity can never outrank the authoritative refold
  (stateVersion bumped for the state-shape change)
- the cold preparation validates the inspected header against the
  enumerated candidate's lifecycle witness; a republished id degrades to
  that child's corrupt diagnostic instead of leaking the new owner's log
- the new projection registration proves HMR disposal; companion notes
  qualify the superseded decision text and record the deliberate
  error-face asymmetry
2026-08-07 00:05:43 +08:00
imccyu
a25d4331d7 feat(subagent): opportunistic projection-cache rung for cold listings
Cold children consult the optional session-projection-cache checkpoint
before paying a preparation recompute: the identity is immutable once
appended, so a cached value is definitive regardless of its watermark.
The cache stays a read-only accelerator (absent service or any rung-two
fault falls through silently; verdicts stay with the authoritative
refold), and the note plus core-data-structures pages describe the
three-rung ladder.
2026-08-07 00:05:42 +08:00
imccyu
0b0b9e4707 fix: address PR #1802 review round
- listChildren reads the session store via strict ctx.get (property proxy
  is caller-scoped), orders candidates branchlessly, narrows the cold-read
  return type, and pins the cost model and store/registry composition gaps
  with tests; per-file coverage restored
- acp-agent and headless-agent compositions mount session-projection; a
  keyless snapshot pins the descriptor-less diagnostic row
- api-proxy cold spec pins header-origin ownership and the legacy
  descriptor-only opt-out
- design note ships as implemented with its English pairing; companion
  notes and core-data-structures pages synced
2026-08-07 00:05:41 +08:00
imccyu
a328fd34d5 feat: subagent list use preparation + projection 2026-08-07 00:05:41 +08:00
kingwl
b30686d634 show subagent usage and active duration 2026-08-02 23:31:24 +08:00
Dudu-0223
431fb4b035 feat(subagent): add explicit child reports 2026-08-02 12:51:10 +08:00
Dudu-0223
a977ef30ee fix(subagent): preserve published run failures 2026-08-02 12:51:09 +08:00
Dudu-0223
03d95d0d84 fix(subagent): keep one-shot labels optional 2026-08-02 12:51:09 +08:00
Dudu-0223
774ee34b9a feat(subagent): catalog one-shot child sessions 2026-08-02 12:51:09 +08:00
Dudu-0223
2e5c439ecf fix(subagent): decouple child listing runtime 2026-08-02 12:51:09 +08:00
Dudu-0223
30b9d0464a fix: address codex review round 3
Recheck the listing cancellation signal after each child inspection settles:
a per-child read failing with a diagnostic-mapped code during an abort
previously bypassed the inspection's own checkpoints, letting a cancelled
single-child scan return a successful result.
2026-08-02 12:51:09 +08:00
Dudu-0223
13631134aa fix: address codex review round 2
- listChildren() checks its cancellation signal before starting the corpus
  trace and immediately after it resolves, so a pre-aborted or trace-time
  abort fails CANCELLED instead of returning a successful empty result.
- Corrected the persistence-repair claims: session query resolves persisted
  candidates through the non-mutating inspect() read, so listing is
  storage-read-only and a torn child tail is surfaced (shorter valid prefix)
  rather than repaired; the Agent Note, README, and JSDoc now state that,
  and the zh pair is re-recorded.
2026-08-02 12:51:09 +08:00
Dudu-0223
4bd98407a9 fix: address codex review round 1
- listChildren() takes an optional AbortSignal and rechecks it after every
  un-signalled session-query await (the cold-resume cooperative-cancellation
  boundary); list_agents forwards exec.signal so the registry's drain of
  started tool bodies cannot block on a slow or large catalog.
- The list_agents description now presents running/complete as a stored-record
  snapshot and defers deliverability to send_message, matching the ownership-
  conflict semantics the service tests pin.
2026-08-02 12:51:09 +08:00
Dudu-0223
4240c7dd7b feat(subagent): durable child catalog and list_agents
Implements the durable-subagent-catalog RFC: SubagentControlService.listChildren()
enumerates a parent's direct continuable children from one sessionQuery trace,
validates each child's sole subagent/descriptor event (now carrying the durable
creation label), and returns one ordered SubagentListEntry[] with per-child
corrupt/unsupported/unavailable diagnostics. The list_agents tool ships as a
separately loadable plugin of dsh-tool-subagent-control requiring sessionQuery
at load; send_message stays usable without it.
2026-08-02 12:51:09 +08:00
Dudu-0223
f4a65a34e6 cleanup(subagent): hide manager-wide continuation drain 2026-08-02 12:51:08 +08:00
Dudu-0223
191c8cd640 fix(acp): scope connection-owned continuation drain 2026-08-02 12:51:08 +08:00
Dudu-0223
a91b20f6be fix(subagent): close continuation lifecycle gaps 2026-08-02 12:51:08 +08:00
Dudu-0223
853f4d5cfb refactor(subagent): drop host-user authority and split lifecycle publication
Remove the host-user continuation capability and the public residency query,
then separate the seam's public event payloads from its internal lifecycle
control interfaces.

`followup()` now takes the exact live direct parent `Agent` instead of a
`SubagentAuthority` union. No production adapter ever supplied user authority,
so the `UserAuthorityGrant` brand token existed only to stop a forged
discriminant from bypassing the direct-parent check — deleting the branch
retires the token, its mint method, and that attack surface together.

Narrowing `parent` from `Agent | undefined` to `Agent` removes three special
cases, including the path where a parentless epoch dispatched its lifecycle
events unscoped. Scoped-versus-global dispatch is now decided by the event, not
by whether a caller happened to have a parent.

`activationState()` had no caller; `ActivationState`, `ActivationObserver`, and
`ContinuationHost` are package-private.

New `src/lifecycle.ts` owns the contained emitter, the one-shot run observer,
and the Activation observer, while `SubagentRunInfo`/`SubagentRunEndInfo` move
to `src/types.ts` beside the other consumer-facing contracts. Those payloads are
public API — dsh-jsonrpc, hooks-claude, and the package invariant all consume
them — whereas the observer is a contract between two in-package collaborators,
so they no longer share a home merely for both being lifecycle-shaped. The
service keeps ownership of the scope carrier: `scopeTarget()` composes the
service's own context filter, so a narrowed stand-in would silently change
scope filtering.

Also drops now-unused dsh-tasks-local and dsh-tool-tasks dev dependencies, and
corrects the README claim that a pre-residency failure emits a terminal edge —
that path only ever rethrew.
2026-08-02 12:51:08 +08:00
Dudu-0223
7428cdf41e fix(subagent): address codex review round 3
- Make host-user authority unforgeable. `{ kind: 'user' }` was a bare
  discriminant, so any plugin holding `ctx.subagents` — including
  model-generated cordis_mount code, which the advanced ACP composition ships
  alongside continuable subagents — could construct it and skip the
  direct-parent check for any known child id. It now carries an opaque grant
  that only SubagentService.userAuthority() mints, which composition hands to
  trusted host adapters; a model-facing tool uses parent authority from its own
  execution context.
- Reconcile a delivery discarded inside its own admission window. An enqueue
  listener that cancels fires the discard before followup() returns, so the
  discard listener could not clear an id it had not seen; submit() retained it
  and residency stayed `running` until an explicit drain.
- Recheck the caller signal after materialization. An abort landing between
  publication and inbox acceptance still submitted the prompt and returned both
  ids; it now rolls the child back.
- Stop promising the model transcript access that no shipped continuable config
  mounts. The tools now state only that a background child does not report back.
- Restate the implemented note as shipped state rather than a proposal, so it
  works as current authority.
2026-08-02 12:51:08 +08:00
Dudu-0223
cbaceb73a9 fix(subagent): address codex review round 2
Round 1 traded one teardown ordering problem for another. The observer now
splits capture from emission, which satisfies both consumers at once:

- Terminal facts are captured while the child is still registered, so consumers
  that resolve it for the child's log and scope still work.
- The edge is emitted only after handle disposal settles, so a rejecting scoped
  cleanup is reported as a failed epoch instead of a successful one.

Also:

- Keep the Activation in the map until disposal settles. Removing it first let a
  racing followup() see no Activation and cold-resume into the still-registered
  agent, and let a concurrent forest drain skip a still-disposing child and
  release its parent first.
- Derive terminal telemetry from this epoch's event suffix rather than the whole
  session, so a cold resume whose prompt is blocked no longer reports the
  previous epoch's answer and turn reason.
- Cancel the ACP bridge's own prompts before awaiting the descendant drain: a
  drain can block on persistence, and the top-level agents must not keep running
  model and tool work for its whole duration.
2026-08-02 12:51:08 +08:00
Dudu-0223
c485b6136d fix(subagent): address codex review round 1
All five findings were real:

- The terminal lifecycle edge derived its stop reason from teardown success, so a
  child that errored, hit its token ceiling, or was cancelled reported as
  completed once its checkpoint and disposal succeeded. It now reads the child's
  own last message turn/end, which is authoritative.
- Live delivery never rechecked the caller signal after authorization yielded, so
  an abort that won before acceptance still enqueued the message and returned an
  id. Admission now re-checks at the boundary that owns the decision.
- Drain flushed before cancelling, letting a running turn keep appending events
  the checkpoint could not cover and letting model work continue through a slow
  flush. It now cancels to quiescence first.
- subagent/end fired after AgentHandle.dispose() unregistered the child, so the
  hooks bridge could not resolve it for the child's cwd and scope. The edge now
  publishes while the child is still registered.
- activationState() read Agent.status alone, which stays idle between an accepted
  waking send and the microtask that admits it, so a synchronous inbox observer
  could see settled with a queued turn. Residency now also counts messages this
  manager admitted but has not seen leave the inbox.
2026-08-02 12:51:08 +08:00
Dudu-0223
3911b7117e test(subagent): reach full continuable coverage and simplify unreachable paths
Adds coverage for the fork-seeded descriptor turn numbering, omitted and declared
descriptor composition fields, a routeless cold resume, and the drain no-op.

Splits materialize's create-versus-resume inputs so the impossible
create-without-meta case disappears, drops the observer's unreachable
pre-residency guard, and annotates the three remaining paths that only a
non-deterministic send-versus-dispose race can reach.
2026-08-02 12:51:08 +08:00
Dudu-0223
bc504195df test(subagent): close continuable coverage and drop unreachable guards
Restores the one-shot settleRun coverage in its own file beside the helper,
covers fork's seed contribution, the post-transfer rollback, the descriptor
model route on cold resume, manager-unload drain, and a failing teardown branch.

Removes three redundant checks the surrounding contracts already own: the
duplicate-Activation and live-id pre-checks (AgentRegistry.enter is the
authoritative collision boundary) and a rollback lifecycle edge that could never
publish because the epoch had no start edge.
2026-08-02 12:51:08 +08:00
Dudu-0223
357f317b4c test(subagent): update service, send_message, and delegation specs
The continuable path has no Task, so send_message reports a queued next turn
and continuable delegation returns only the durable child id. Pins that a
follow-up queues behind an open turn rather than steering it, and that a
non-parent caller is rejected.

Also makes startContinuable/followup reject rather than throw synchronously
when continuation services are absent, so callers have one failure mode.
2026-08-02 12:51:08 +08:00
Dudu-0223
26a117f842 feat(subagent): activation-based continuable subagents (source)
Replace the Task-backed continuation manager with one durable Session plus at
most one process-local Activation — a residency epoch for a reconstructed child
Agent, not a request, result, cancellation, or Task boundary. The manager owns
activation admission, authority, the live ownership graph, cold resume, and
child-first disposal; the Agent inbox is the only turn FIFO.

- startContinuable() is async and returns { childId, messageId } at inbox
  acceptance; followup() takes a SubagentAuthority and returns AgentMessageId.
- SubagentProvider.resume?(), SubagentProviderResumeRequest, SubagentRun.steer?(),
  SubagentProviderStartRequest and SubagentContinuation are deleted;
  prepareContinuable?() is the continuable-creation capability.
- Cold resume calls ctx.agents.resume() from the manager through a private
  activation-owner scope, never dispatching through a provider.
- Extract shared child composition, descriptor seeding, depth accounting, and
  one-shot run settlement so the manager and one-shot driver keep one home
  per fact.

Tests and docs follow in subsequent commits.
2026-08-02 12:51:08 +08:00
Dudu-0223
264bc41a13 fix(subagent): preserve ordinary start requests 2026-08-02 04:34:17 +08:00
Tianyi Cui
f14121a4c2 refactor(subagent): narrow continuation interface 2026-08-02 04:34:17 +08:00
Dudu-0223
644bf00b86 fix(subagent): cancel pending live delivery 2026-08-02 04:34:17 +08:00
Dudu-0223
112e25bb05 refactor(subagent): unify service errors 2026-08-02 04:34:17 +08:00
Dudu
88f913a9ae refactor(subagent): merge continuation control service 2026-08-02 04:34:16 +08:00
Dudu-0223
99a778d63f feat(subagent): continuable background subagents
Implement the continuable background subagents RFC: a durable child
session with a series of Task-backed activations, each disposing its
run before the Task settles.

- dsh-subagent: rename SubagentRun.sendMessage to strict steer, drop
  run-level resume, add SubagentProvider.resume dispatch via
  SubagentService.resume, the continuation start field, and the
  versioned model-hidden subagent/descriptor session event.
- dsh-subagent-inprocess/-spawn/-fork: publish the control-allocated
  child id, append the descriptor inside the initial turn, implement
  cold resume from the child's own transcript under the live parent
  scope, and strict running-only steer.
- dsh-subagent-control (new): SubagentControlService owning stable
  child ids, descriptor snapshot/fold/authorization, Task-backed
  activation with settle-then-dispose ordering, the process-local
  active-run association, and steer-or-resume sendMessage routing.
- dsh-tool-subagent: background route branches on the provider's
  resume capability (continuable via the control service; one-shot
  task for ACP), returning both child and task ids.
- dsh-tool-subagent-control (new): the globally named send_message
  tool rendering steered/started routes.

Keyless coverage spans Task ownership and disposal ordering, running
delivery, cold follow-up, descriptor rejection and rollback, known-id
reconstruction, kill during lookup, admission races, and a new
subagent-continuable ACP snapshot scenario.
2026-08-02 04:34:15 +08:00
Turtle
95a995968b refactor: migrate linting to Oxlint 2026-07-29 14:32:11 +08:00
Tianyi Cui
87a249106f Merge origin/master: port the SDK stack onto the subprocess seam
Master's #660 replaced dsh-subagent-subprocess with the dsh-subprocess
capability seam (ctx.subprocess + scrubbedParentEnv, tree-scoped teardown)
and moved subagent-acp onto it. Convergence for this branch's packages:

- The shared out-of-process provider vocabulary this branch had grown in
  the deleted library (NO_START_CAPABILITIES, assertPositiveFinite, cwd
  resolution, settleRunResult, subprocessRunHandle) moves into the subagent
  seam package as out-of-process.ts — it enforces subagent-seam contracts,
  not process mechanics, and both out-of-process backends now import it
  from there (subagent-acp keeps master's shape otherwise).
- subagent-sdk spawns THROUGH the SDK client (the subprocess README's
  documented exception for SDK-managed transports) and now applies the
  seam's scrubbedParentEnv() + explicit-env merge in place of the deleted
  buildChildEnv.
- sdk-client inlines the EOF→SIGTERM→SIGKILL ladder as private helpers (it
  runs outside any harness context, so it cannot ride ctx.subprocess).
- The child harness fixture gains the now-required dsh-subprocess-local
  entry for bash-local; the fixture cordis.yml keeps exercising the
  shipped provider default.
2026-07-27 21:52:40 +08:00
Tianyi Cui
8500974fd4 feat: unify JSON value schema DSL 2026-07-21 01:11:55 +08:00
Tianyi Cui
ff6359a0cd docs(subagent): close review contract gaps 2026-07-20 18:10:31 +08:00
Yichen Jiang
0d00106fa8 fix(session): persist the delegation depth in the session header
A subagent child's recursion depth lived only in runtime AgentOptions,
so a persisted child came back from resume counted as top-level and
maxDepth stopped binding after every restart. Add
SessionHeader.delegationDepth, round-trip it through the JSONL and
SQLite backends (SQLite schema v5), restore it on agent-loop resume,
and write it when the in-process backends create a child. The seam now
owns the shared depth vocabulary (delegationDepthOf): the persisted
header is authoritative and monotone — runtime options may deepen it
but never lower it.
2026-07-19 17:20:36 +08:00