/** * Tool registry and execution pipeline. Plugins register tools; the registry * feeds schemas into the system prompt, and `execute()` dispatches each call * through `tools/pre-execute` (the extensible allow/deny gate) → monotonic * registered guards → `tools/execute` (an around-dispatch wrapper for * timeout/retry/metrics plugins) → `tools/post-execute` (inspect/replace the * result, attach context) → the observe-only `tools/result` notification. * * The registry also owns HOW its tools are presented to the model — its * `mode` config: `'native'` (every tool as a wire function definition, * today's behavior and the default), `'code'` (the registry's canonical wire * contribution is one tool, `run_code`, plus a generated TypeScript SDK prompt section), or * `'both'`. See `code-mode.ts` (the tool + dispatch bridge) and * `ts-types.ts` (the SDK codegen); design in the Code Mode RFC. * * @module @deepseek-ai/dsh-tools */ import { Context, Service } from 'cordis' import z from 'schemastery' import { scopeOf, scopeTarget } from '@deepseek-ai/dsh-scope' import type { ScopeKey, Scoped } from '@deepseek-ai/dsh-scope' import type { CallId, ContentBlock, ToolSchema } from '@deepseek-ai/dsh-llm' import { assertNever, deepFreeze, HarnessError } from '@deepseek-ai/dsh-llm' import type { Agent, HookContext } from '@deepseek-ai/dsh-agent' import { snapshotJsonValue } from '@deepseek-ai/dsh-session' import type { ToolProviderResult } from '@deepseek-ai/dsh-system-prompt' import type { CodeRuntime } from '@deepseek-ai/dsh-code-runtime' // Type-only: makes `ctx.get('approval')` resolve to the ApprovalService // augmentation. The seam stays optional at runtime — see `serviceAsk`. import type {} from '@deepseek-ai/dsh-user-approval' import type { ToolCallView, ToolResultView } from './presentation.ts' import { createRunCodeTool, RUN_CODE_NAME, SDK_SECTION_ORDER } from './code-mode.ts' import { renderToolsSdk } from './ts-types.ts' export { defineTool, schemaSpecToJsonSchema, validateArgs, ToolArgsError, type SchemaSpec, type SchemaProp, type SchemaType, type InferArgs, type DefineToolOptions, type JsonSchemaObject, } from './schema.ts' export { assertSupportedOutputSchema, validateStructuredValue, OutputSchemaError, type StructuredOutputSchema, type StructuredSchemaNode, type StructuredSchemaType, type StructuredScalar, } from './json-schema.ts' export { CodeRunFailedError, RUN_CODE_NAME } from './code-mode.ts' export { jsonSchemaToTs, renderToolsSdk } from './ts-types.ts' // The render-intent vocabulary a tool declares via `presentCall`/`presentResult` // lives in its own UI-facing module; re-export it so `@deepseek-ai/dsh-tools` // stays the single public surface for consumers (producers + the ACP bridge). export type { ToolCallKind, FileLocation, FileDiff, ToolCallView, GenericCallView, TerminalCallView, DiffCallView, ToolResultView, GenericResultView, TerminalResultView, DiffResultView, } from './presentation.ts' declare module 'cordis' { interface Context { tools: ToolRegistry } interface Events { /** * Waterfall BEFORE a tool runs — the gate where sandbox, permission, and * hook plugins allow or deny a call (Claude Code's `PreToolUse`). Listeners * receive `(exec, next)`: call `next()` to delegate to the default (allow), * or return a {@link PreToolDecision} without calling `next()` to * short-circuit. A `deny` skips dispatch and yields an `isError` result; the * tool body never runs. Input rewrite is deliberately NOT offered here (see * {@link PreToolDecision}); `ask` is serviced by the `ctx.approval` seam * when one is mounted, and degrades to deny otherwise. * Scope-filtered dispatch (`@deepseek-ai/dsh-scope`) keys the carrier by `exec.agent`: a * listener registered through `agent.ctx` fires only for that agent's * calls, while a plain plugin listener fires for every call (including * agent-less ones, which dispatch subject-less). * @param exec - the pending call (name, parsed arguments, caller agent). * @mode waterfall */ 'tools/pre-execute'(this: Scoped, exec: ToolExecution, next: () => Promise): Promise /** * Around-dispatch waterfall wrapping the registry's core tool dispatch, * between the `tools/pre-execute` gate and the `tools/post-execute` seam. A * listener receives `(exec, next)`: call `next()` to delegate to dispatch * (returning its {@link ToolExecutionResult}, optionally wrapped), or return a * replacement result without calling `next()` to short-circuit dispatch. The * base `next()` IS the dispatch-with-normalization thunk — a thrown tool (or * unknown tool) is already normalized to an `isError` result by the time a * listener's `await next()` returns, so a wrapper never sees a raw throw from * the tool body. This is the seam a timeout/retry/metrics plugin wraps: it can * set or replace the one mutable field, `exec.signal` (e.g. with a per-call * deadline), BEFORE `next()`, restore/delete it afterward, and inspect the result AFTER. Call identity * (`token`, `callId`, `name`, `arguments`, `agent`, and `parent`) is immutable throughout the * pipeline so a wrapper cannot change which tool and scope the pipeline * accepted. (Cordis `next()` ignores passed arguments and re-invokes * downstream with the shared payload, so a wrapper changes `exec.signal` in * place rather than passing a new object to `next()`.) * Multiple listeners compose by registration order — an outer one wraps the * inner ones plus dispatch. * Scope-filtered dispatch (`@deepseek-ai/dsh-scope`): the carrier is keyed by * `exec.agent` — a listener registered through `agent.ctx` wraps only that * agent's calls; a plain plugin listener wraps every call (including * agent-less ones, which dispatch subject-less). * @param exec - the allowed call about to dispatch (name, parsed arguments, caller agent, signal). * @mode waterfall */ 'tools/execute'(this: Scoped, exec: ToolExecution, next: () => Promise): Promise /** * Waterfall AFTER a tool runs — where hook plugins inspect the result and * accept it (optionally REPLACING the model-facing content, and/or attaching * `additionalContext` for the next request) or block it with corrective * `feedback` (Claude Code's `PostToolUse`). Listeners receive * `(exec, result, next)`: call `next()` to delegate to the default (accept * unchanged), or return a {@link PostToolDecision} to override. Core tool * dispatch runs earlier as the base `next()` of the `tools/execute` * waterfall, all inside `execute`'s outer try/catch (and the tool body keeps * its own inner try/catch, so a thrown tool still reaches `post-execute` as an * `isError` result). * Scope-filtered dispatch (`@deepseek-ai/dsh-scope`): the carrier is keyed by * `exec.agent` — a listener registered through `agent.ctx` fires only for * that agent's calls; a plain plugin listener fires for every call * (including agent-less ones, which dispatch subject-less). * @param exec - the call that just ran (name, parsed arguments, caller agent). * @param result - the dispatch outcome a listener may accept, replace, or block. * @mode waterfall */ 'tools/post-execute'(this: Scoped, exec: ToolExecution, result: Readonly, next: () => Promise): Promise /** * Synchronous notification of the authoritative FINAL tool outcome, after the * complete pre/execute/post pipeline, final lossless-JSON validation, and * outer error normalization. * Unlike the three waterfalls, this seam cannot transform the result: each * listener receives the now-frozen execution object and a deep-frozen result * snapshot; listener failures are contained and logged, and * {@link ToolRegistry.execute} still returns the outcome. * Scope-filtered dispatch (`@deepseek-ai/dsh-scope`): keyed by * `exec.agent`, using the same carrier as the pipeline. * @param exec - the execution object that traversed the pipeline. * @param result - a deep-frozen snapshot of the final returned result. * @mode emit */ 'tools/result'(this: Scoped, exec: Readonly, result: Readonly): undefined /** * A tool was registered or unregistered, or a scoped restriction changed * (the available tool set changed — possibly for one scope only). An * UNFILTERED registry-subject notification, deliberately not scope-filtered * dispatch: a global change concerns every agent's next assembly, so a * scoped listener subscribing here sees every change, not just its own * scope's. * @mode emit */ 'tools/change'(): void } } // TODO(review): revisit these shapes when concurrency metadata becomes useful // (for example, a read-only hint that would permit safe parallel execution). /** * What a tool's `execute` returns. The bare {@link ContentBlock}`[]` form is the * common case (model-facing content only); the object form additionally attaches * a tool-private `meta` presentation payload that the registry threads onto the * `tool/result` session event and hands back to the tool's `presentResult`. * `meta` is opaque to the core (`unknown` — the tool owns and narrows its shape), * and MUST be JSON-serializable: it persists on the durable log (the session * enforces this at `append`), so replay reproduces the card. */ export type ToolExecuteReturn = ContentBlock[] | { content: ContentBlock[]; meta?: unknown } /** A registered tool: its schema plus the execution function. */ export interface ToolDefinition extends ToolSchema { execute(args: unknown, exec: ToolExecution): Promise /** * Cooperative tool-call timeout budget in milliseconds. Omit for no deadline. * Enforced by `@deepseek-ai/dsh-timeout-policy` (a `tools/execute` wrapper); it * is NEVER sent to the model — `schemas()` whitelists only name/description/ * parameters. Declaring it asserts this tool forwards `exec.signal` to a * cooperative implementation that can reach quiescence when the signal aborts. */ timeoutMs?: number /** * Optional: how to present the PENDING state of one call in a UI, derived from * the call's `args` (parsed arguments, `unknown` — the tool validates/narrows * its own input). Returns a {@link ToolCallView} (a `card`-tagged render intent), * or `undefined` (or omit the method) to fall back to a generic presentation * (title = tool name, raw args as input). Pure and side-effect-free: a UI may * call it during live streaming AND a session-log replay, so it must depend * only on `args`. */ presentCall?(args: unknown): ToolCallView | undefined /** * Optional: how to present the COMPLETED state, given the same `args` and the * `result` (`execute`'s content + whether it errored). Returns a * {@link ToolResultView}, or `undefined` (or omit the method) to keep the * pending title and render the raw result content. Pure and side-effect-free * for the same replay reason. */ presentResult?(args: unknown, result: ToolResult): ToolResultView | undefined } /** The completed outcome handed to {@link ToolDefinition.presentResult}. */ export interface ToolResult { /** The model-facing content `execute` returned (or the error text on failure). */ content: ContentBlock[] /** Whether the call failed. */ isError: boolean /** * The tool-private presentation payload the tool attached from `execute` (via * the object return form), threaded verbatim from the `tool/result` event. * Opaque (`unknown`); the tool narrows it back to its own shape. Absent when * the tool attached none. */ meta?: unknown } declare const toolExecutionTokenBrand: unique symbol /** * Opaque identity for one trip through the tool pipeline. Nested * transports carry the enclosing execution's token instead of its live object, * so observe-only result listeners can correlate calls without gaining a * mutation path into an outer around-dispatch wrapper. */ export type ToolExecutionToken = symbol & { readonly [toolExecutionTokenBrand]: true } /** * Caller-supplied description of one tool call. {@link ToolRegistry.execute} * adds the registry-owned token to form a pipeline {@link ToolExecution}; * callers do not choose that token. */ export interface ToolExecutionInput { readonly callId: CallId readonly name: string /** Losslessly JSON-serializable parsed arguments (tools validate their own schema). */ readonly arguments: unknown /** The agent on whose behalf the call runs (set by the agent loop). */ readonly agent?: Agent /** * Opaque token of the enclosing transport execution, when one exists. Code * Mode sets this on SDK sub-dispatches so commit-style observers can wait for * the outer `run_code` outcome without receiving its live mutable execution. */ readonly parent?: ToolExecutionToken signal?: AbortSignal } /** * One pending tool call inside the registry pipeline. Parsed arguments cross * one lossless-JSON materialization boundary before policy and are deep-frozen; * call identity and the registry-assigned {@link token} are readonly. An * around-dispatch wrapper may set, replace, or remove `signal`. The registry * freezes the complete object before `tools/result` observers run. */ export interface ToolExecution extends ToolExecutionInput { /** Registry-assigned identity shared with nested calls only as their opaque `parent` token. */ readonly token: ToolExecutionToken } /** Structured error metadata for a failed tool call (alongside the model-facing text). */ export interface ToolErrorInfo { name: string code: string } /** * Thrown (internally) when the model requests a tool that isn't registered. * Extends {@link HarnessError} (`code: 'UNKNOWN_TOOL'`) so an unknown-tool * failure is as routable as a tool-thrown one — retry/sandbox/replay code can * distinguish it from a tool body's own error. */ export class ToolNotFoundError extends HarnessError { constructor(public readonly toolName: string) { super(`unknown tool "${toolName}"`, 'UNKNOWN_TOOL') this.name = 'ToolNotFoundError' } } /** The outcome of one tool call. */ export interface ToolExecutionResult { callId: CallId content: ContentBlock[] isError: boolean /** * Set when the call failed with a {@link HarnessError}: machine-routable * `{ name, code }` for retry/sandbox plugins and replay. The model-facing * text in `content` is always present; this is extra structure for code. */ error?: ToolErrorInfo /** * Extra model-facing context a `tools/post-execute` listener attached for the * NEXT request (Claude Code's PostToolUse `additionalContext`). It is NOT part * of this call's `content` — `content`/`feedback` shape the tool RESULT, but * `additionalContext` is a SEPARATE `context/message`. A step can carry * multiple tool calls, so the loop BUFFERS every call's `additionalContext` * and appends them only AFTER all `tool/result`s for the step, keeping * tool-call/result adjacency intact. Carried on the result purely to ferry it * from `execute()` up to the loop's per-step buffer. */ additionalContext?: HookContext /** * The tool-private presentation payload from a successful `execute` (the object * return form). Threaded onto the `tool/result` session event and back into * {@link ToolResult} for `presentResult`. Opaque (`unknown`); absent when the * tool attached none or the call failed. */ meta?: unknown } /** * The decision a `tools/pre-execute` listener returns for one pending call. * Maps onto Claude Code's `PreToolUse` `permissionDecision`. * * - `allow` proceeds to dispatch. (Input rewrite — changing `exec.arguments` — * is deliberately NOT offered: `tool/call` and `assistant/message` are logged * BEFORE execution and live consumers, e.g. the ACP bridge and `dsh-tool-bash` * presentation, read the pre-execution arguments, so an execution-only rewrite * would desync the UI from what RAN. That consistency redesign is its own * `proposed` RFC; `TODO(pre-tool-input-rewrite)` anchors it at the call site.) * - `deny` skips dispatch; the loop records an `isError` result carrying `reason`. * - `ask` is the permission-prompt intent: serviced as a one-shot decision by * the `ctx.approval` seam when one is mounted (`allowed-once` proceeds to * dispatch; every other outcome denies), degrading to `deny` when none is. */ export type PreToolDecision = | { kind: 'allow' } | { kind: 'deny'; reason: string } | { kind: 'ask'; reason?: string } /** * The decision a `tools/post-execute` listener returns for one finished call. * Maps onto Claude Code's `PostToolUse` decision. * * - `accept` keeps the call successful; optional `content` REPLACES the * model-facing result (clean: `tool/result` is logged AFTER `execute()` * returns, so a replaced result is the single source of truth for both derived * history and UI). Optional `additionalContext` rides to the next request. * - `block` turns the call into an `isError` result whose content is the * corrective `feedback` (the model is told the call was rejected and why), * optionally also attaching `additionalContext`. */ export type PostToolDecision = | { kind: 'accept'; content?: ContentBlock[]; additionalContext?: HookContext } | { kind: 'block'; feedback: ContentBlock[]; additionalContext?: HookContext } /** * Best-effort human-readable message from an arbitrary thrown value: Error * instances use `.message`; non-Error objects with a string `message` * property (e.g. `throw { message: 'denied' }`) use it too; everything else * is stringified. */ function errorMessage(error: unknown): string { try { if (error instanceof Error) return error.message if (typeof error === 'object' && error !== null && 'message' in error && typeof error.message === 'string') { return error.message } return String(error) } catch { // A hostile thrown value can trap `instanceof`, property access, or string // coercion. Error normalization is the outermost safety boundary, so its // fallback must itself be total. return '' } } /** Structured `{ name, code }` for a thrown HarnessError, else undefined. */ function errorInfo(error: unknown): ToolErrorInfo | undefined { try { return error instanceof HarnessError ? { name: error.name, code: error.code } : undefined } catch { return undefined } } /** How the registry presents its tools to the model (see {@link Config.mode}). */ export type ToolPresentationMode = 'native' | 'code' | 'both' /** Plugin config: how the registered tools are presented to the model. */ export interface Config { /** * The presentation mode. `'native'` (the default) contributes every * visible end capability as a native wire function definition. Under * `'code'` this registry contributes exactly ONE wire tool, * `run_code`, plus the generated `tools:sdk` prompt section declaring every other tool as a * TypeScript API the program calls. `'both'` contributes every native * definition AND `run_code` + the SDK section. Non-native modes require a * loaded `ctx.codeRuntime` whose `language` is `'typescript'` — a missing * or mismatched runtime rejects every prompt assembly with an actionable * error (misconfiguration fails loud, before any model request). A * configured `systemPrompt.toolOrder` naming native tools likewise rejects * every assembly under `'code'` (those names are no longer contributed) — * a deployment switching modes updates its order config or drops it. */ mode?: ToolPresentationMode } /** * A per-scope restriction over the GLOBAL tool surface, registered via * {@link ToolRegistry.restrict}. `allow` keeps only the listed global tools; * `deny` removes the listed ones; both present = allow first, then deny. * Restrictions never touch scoped registrations — a tool registered through * the same scope is merged after the global filter (which is what keeps e.g. a * structured-output capture tool alive under an allow-list). The readonly * filter values compile to private sets at registration, but resolution uses the live global registry: * a later global name passes a deny-only filter unless explicitly denied and * fails an allow-list unless explicitly allowed. The * reserved `run_code` presentation transport is likewise outside capability * filtering, and naming it explicitly is rejected. Multiple restrictions on * one scope compose by intersection: every one must admit. */ export interface ToolRestriction { /** Global tool names that stay visible; everything else is removed. */ readonly allow?: readonly string[] /** Global tool names removed from visibility. */ readonly deny?: readonly string[] } /** One restriction compiled at registration for repeated live-global lookup. */ interface CompiledToolRestriction { readonly allow?: ReadonlySet readonly deny?: ReadonlySet } /** One scope's complete registry view, derived in a single layer traversal. */ interface ToolView { /** Visible definitions after restrictions, scoped shadowing, and transport insertion. */ readonly visible: ReadonlyMap /** Pre-restriction capability names used by prompt-order validation. */ readonly knownNames: ReadonlySet /** Current global names that a scoped restriction may name. */ readonly restrictableNames: ReadonlySet } /** * A monotonic execution guard evaluated after every `tools/pre-execute` * listener and before the tool body. Returning a reason denies the call; * returning `undefined` leaves it unchanged. Because guards have no allow * result, listener ordering cannot turn a denial back into permission. * @param execution - the identity-protected call after extensible pre-execute policy completed. * @returns a final denial reason, or `undefined` to leave the call allowed. */ export type ToolGuard = (execution: Readonly) => string | undefined /** One guard registration; the wrapper preserves independent duplicate registrations. */ interface ToolGuardRegistration { guard: ToolGuard } /** * Tool registry (`ctx.tools`): tool plugins register definitions; the agent * loop executes calls through the `tools/pre-execute` → guards → * `tools/execute` → `tools/post-execute` → `tools/result` pipeline. The * registry contributes its schemas into the system-prompt assembly — WHICH * schemas is governed by its `mode` config * (see {@link Config.mode}); under a non-native mode it also owns the reserved * `run_code` presentation transport and the `tools:sdk` prompt section. * * Two registration layers (`@deepseek-ai/dsh-scope`): a registration through a * plain plugin context is GLOBAL (visible to every agent); one through a * scoped context (`agent.ctx`) is filed in that scope's layer — visible to * that agent alone, disposed with the scope, and SHADOWING a global tool of * the same name for that agent (most-specific-wins; within one layer a * duplicate name still throws). {@link restrict} masks the global layer per * scope. One private visibility resolver feeds the registry's prompt * contribution, {@link get}, and {@link execute} — and, under a non-native * mode, the SDK section and `run_code`'s bindings — so those registry-owned * presentation and dispatch paths agree. An expert `system-prompt/assemble` * listener may deliberately replace the final wire composition and owns any * resulting divergence. */ export class ToolRegistry extends Service { static inject = ['systemPrompt'] static Config: z = z.object({ mode: z.union(['native', 'code', 'both'] as const).default('native'), }) private global = new Map() private scoped = new Map>() /** Compiled restriction filters, per scope (see {@link restrict}). */ private restrictions = new Map() /** Monotonic post-policy guards, split into global and per-agent layers. */ private globalGuards = new Set() private scopedGuards = new Map>() private readonly mode: ToolPresentationMode /** Reserved presentation transport, kept outside the filterable registration layers. */ private readonly codeTransport: ToolDefinition | undefined constructor(ctx: Context, config: Config = {}) { super(ctx, 'tools') // The schema already defaulted an omitted mode; the ?? narrows the // optional-input type for direct (non-Loader) construction in tests. this.mode = config.mode ?? 'native' // `run_code` is presentation infrastructure, not an end capability. It // therefore does not enter the global layer: per-agent restrictions must // not remove it, and a scoped registration must not shadow it. The // visibility resolver appends this reserved definition after resolving // the filterable global/scoped capability layers. this.codeTransport = this.mode === 'native' ? undefined : createRunCodeTool(this, () => this.requireCodeRuntime()) ctx.systemPrompt.tools(context => this.wireSchemas(context.scope)) if (this.mode !== 'native') { ctx.systemPrompt.section({ name: 'tools:sdk', order: SDK_SECTION_ORDER, // A lazy thunk over the live registry, per assembly CONTEXT: // regenerated at each assembly over the CALLING SCOPE's visible set // (scoped tools join, restricted globals vanish — the SDK declares // exactly what that agent's programs can call), in lexicographic // tool order, so an unchanged tool set renders byte-identical text // (prefix-cache-friendly) and a mid-session registration surfaces // exactly like a native-mode tool change. text: (context) => { this.requireCodeRuntime() return renderToolsSdk(this.schemas(context.scope).filter(schema => schema.name !== RUN_CODE_NAME)) }, }) } } /** * The registry's contribution to the wire tool list, per {@link Config.mode}, * as ONE SCOPE sees it (scoped layer joins, shadowing and restrictions * applied — {@link schemas}). Because `PromptAssembly.tools` is what the * loop's request header snapshots, the mode's collapse is logged and * reconstructable for free. Under a non-native mode this is also the loud * misconfiguration gate: no usable code runtime → every assembly rejects * before any model request. * * The `knownNames` universe distinguishes the two ways a tool can be off * the wire: a per-scope RESTRICTION is runtime state, so `knownNames` stays * pre-restriction and a restricted-away tool in `toolOrder` is a normal * absence — while the MODE collapse is deployment config, so under * `mode: 'code'` the universe is `[run_code]` and a `toolOrder` naming a * native tool is dead configuration that fails every assembly loud. Under * `mode: 'both'`, the provider adds the reserved transport to the * capability-only known-name universe for `toolOrder` validation. */ private wireSchemas(scope?: ScopeKey): ToolProviderResult { const view = this.view(scope) const schemas = [...view.visible.values()].map(definition => this.schemaOf(definition, false)) if (this.mode === 'native') { return { schemas, knownNames: [...view.knownNames] } } this.requireCodeRuntime() if (this.mode === 'code') { return { schemas: schemas.filter(schema => schema.name === RUN_CODE_NAME), knownNames: [RUN_CODE_NAME], } } return { schemas, knownNames: [...view.knownNames, RUN_CODE_NAME] } } /** * Resolve the code runtime or throw the actionable misconfiguration error. * Read at use time (assembly / run_code execution), NOT via static * `inject`: an inject entry would hold `ctx.tools` — and every tool plugin * behind it — hostage to a code runtime existing even under `mode: * 'native'` (the loop's optional-backend idiom, same as * `sessionPersistence`). */ private requireCodeRuntime(): CodeRuntime { const runtime = this.ctx.get('codeRuntime') if (!runtime) { throw new Error(`dsh-tools: mode "${this.mode}" requires a code runtime — load a ctx.codeRuntime implementation (e.g. @deepseek-ai/dsh-code-runtime-worker) or set tools mode to "native"`) } if (runtime.language !== 'typescript') { throw new Error(`dsh-tools: mode "${this.mode}" generates a TypeScript SDK, but the loaded code runtime's language is "${runtime.language}"`) } return runtime } /** * Register a tool. The layer is decided by the CALLING context: a plain * plugin context registers globally; a scoped context (`agent.ctx`) * registers into that scope's layer — visible to that agent alone, disposed * with the scope, and shadowing a same-named global tool for that agent. * Throws if the SAME layer already has the name (cross-layer name twins are * the shadowing feature, not an error; the global-duplicate message names * `agent.ctx` as the per-agent alternative), or if a non-native mode reserves * the `run_code` name for its presentation transport. The visible schema set * flows into prompt assembly automatically. Definitions are trusted typed * same-process contributions; JSON materialization happens when the schema or * result reaches its model/log boundary. Emits `tools/change` on * register/unregister. * @param definition - the tool's schema plus its execute (and optional * presentation) functions. * @returns the disposer that unregisters the tool. The exact * Cordis effect disposer (single-shot): composite (generator) effects may * yield it directly — exact identity nests the teardown in order. */ register(definition: ToolDefinition): () => void { const scope = scopeOf(this.ctx) const name = definition.name const timeoutMs = definition.timeoutMs if (timeoutMs !== undefined && (!Number.isFinite(timeoutMs) || timeoutMs <= 0)) { throw new TypeError(`tool "${name}" timeoutMs must be a positive finite number`) } if (this.codeTransport !== undefined && name === RUN_CODE_NAME) { throw new Error(`tool name "${RUN_CODE_NAME}" is reserved for the Code Mode presentation transport and cannot be registered or shadowed`) } const dispose = this.ctx.effect(function* (this: ToolRegistry) { const layer = scope === undefined ? this.global : this.layerFor(scope) if (layer.has(name)) { throw new Error(scope === undefined ? `tool "${name}" is already registered (for a per-agent variant, register through that agent's \`agent.ctx\` instead)` : `tool "${name}" is already registered in this scope`) } layer.set(name, definition) // Yield the rollback BEFORE emitting `tools/change`: a generator effect // collects each yielded disposer before the next step runs, so a throwing // `tools/change` listener removes the tool instead of leaking it (a leak // would wedge the duplicate-name check until restart). The duplicate // throw above fires before any mutation — it leaks nothing. yield () => { layer.delete(name) // An emptied scope layer is dropped so a disposed scope leaves no // residue keyed by its (dead) key. if (scope !== undefined && layer.size === 0) this.scoped.delete(scope) this.ctx.emit('tools/change') } this.ctx.emit('tools/change') }.bind(this), 'tools.register()') // The EXACT cordis effect disposer, not a wrapper: a composite (generator) // effect that owns a teardown ORDER must be able to yield THIS function — // cordis nests a disposer out of the fiber's concurrent sibling list by // exact function identity, so a wrapper would silently break the nesting // (the agents.register() lesson). Cleanup is synchronous because this // registration installs only synchronous state and notifications. // eslint-disable-next-line @typescript-eslint/no-misused-promises -- synchronous cleanup; direct return preserves disposer identity return dispose } /** * Restrict the GLOBAL tool surface for the calling scope. Must be called * through a scoped context (`agent.ctx`) — restricting "everyone" is not a * thing (throw), and an empty filter (neither `allow` nor `deny`) is a no-op * that can only be a bug (throw — the materialized-empty-config trap). * Validates every listed name against the CURRENT global end-capability * universe and throws on an unknown or scope-local name (fail loud * beats a typo silently filtering nothing) — register restrictions after the * global tools they mask exist (the agent-creation `setup` window satisfies * this). A non-native mode's reserved `run_code` presentation transport is * not a filterable capability; naming it explicitly throws, while omitting * it from an allow-list cannot remove it. The readonly arrays are compiled to * private sets at registration. Resolution still uses the live global registry, so a later * global name passes a deny-only filter unless named and fails an allow-list * unless named. Multiple restrictions compose by intersection. Scoped * registrations are merged after restrictions and therefore remain visible. * Disposed with the calling fiber (revocable independently); emits * `tools/change`. * @param filter - global-surface mask: `allow` (keep only) and/or `deny` (remove). * @returns the disposer that lifts this restriction. The exact * Cordis effect disposer (single-shot): composite (generator) effects may * yield it directly — exact identity nests the teardown in order. */ restrict(filter: ToolRestriction): () => void { const scope = scopeOf(this.ctx) if (scope === undefined) { throw new Error('tools.restrict() requires a scoped context (agent.ctx): a context-global restriction would mask every agent — deny the tool for the intended agent instead') } const allow = filter.allow const deny = filter.deny if (allow === undefined && deny === undefined) { throw new Error('tools.restrict({}) is a no-op: pass `allow` and/or `deny` (an empty filter is almost always a materialized-empty-config bug)') } const compiled: CompiledToolRestriction = { ...allow !== undefined ? { allow: new Set(allow) } : {}, ...deny !== undefined ? { deny: new Set(deny) } : {}, } if (this.codeTransport !== undefined && [...allow ?? [], ...deny ?? []].includes(RUN_CODE_NAME)) { throw new Error(`tools.restrict() cannot name reserved Code Mode presentation transport "${RUN_CODE_NAME}"; restrict end-capability tools instead`) } const known = this.view(scope).restrictableNames const unknown = [...allow ?? [], ...deny ?? []].filter(name => !known.has(name)) if (unknown.length > 0) { throw new Error(`tools.restrict() names unknown global tool${unknown.length > 1 ? 's' : ''} ${unknown.map(n => `"${n}"`).join(', ')}; known global tools: ${[...known].sort().join(', ') || '(none)'}`) } const dispose = this.ctx.effect(function* (this: ToolRegistry) { const list = this.restrictions.get(scope) ?? [] this.restrictions.set(scope, list) list.push(compiled) yield () => { const index = list.indexOf(compiled) /* v8 ignore next 3 -- defensive: the compiled restriction was pushed, so indexOf is guaranteed >= 0 */ if (index >= 0) list.splice(index, 1) if (list.length === 0) this.restrictions.delete(scope) this.ctx.emit('tools/change') } this.ctx.emit('tools/change') }.bind(this), 'tools.restrict()') // The EXACT cordis effect disposer, not a wrapper: a composite (generator) // effect that owns a teardown ORDER must be able to yield THIS function — // cordis nests a disposer out of the fiber's concurrent sibling list by // exact function identity, so a wrapper would silently break the nesting // (the agents.register() lesson). Cleanup is synchronous because this // registration installs only synchronous state and notifications. // eslint-disable-next-line @typescript-eslint/no-misused-promises -- synchronous cleanup; direct return preserves disposer identity return dispose } /** * Register a monotonic guard after the extensible `tools/pre-execute` * waterfall. A plain-context guard applies globally; one registered through * `agent.ctx` applies only to that agent. Any matching guard may deny by * returning a reason, while no guard can force-allow a call another guard * denied. The exact effect disposer is returned for ordered ownership and * HMR cleanup. * @param guard - synchronous check; a returned string denies the execution. * @returns the exact disposer that unregisters the guard. */ guard(guard: ToolGuard): () => void { const scope = scopeOf(this.ctx) const registration = { guard } const dispose = this.ctx.effect(function* (this: ToolRegistry) { const layer = scope === undefined ? this.globalGuards : this.guardLayerFor(scope) layer.add(registration) yield () => { layer.delete(registration) if (scope !== undefined && layer.size === 0) this.scopedGuards.delete(scope) } }.bind(this), 'tools.guard()') // eslint-disable-next-line @typescript-eslint/no-misused-promises -- synchronous cleanup; direct return preserves disposer identity return dispose } /** The (created-on-demand) scoped layer for `scope`. */ private layerFor(scope: ScopeKey): Map { let layer = this.scoped.get(scope) if (!layer) { layer = new Map() this.scoped.set(scope, layer) } return layer } /** Get or create the guard layer for one agent scope. */ private guardLayerFor(scope: ScopeKey): Set { let layer = this.scopedGuards.get(scope) if (layer === undefined) { layer = new Set() this.scopedGuards.set(scope, layer) } return layer } /** First monotonic denial from the global then matching scoped guard layers. */ private guardReason(exec: ToolExecution): string | undefined { for (const { guard } of this.globalGuards) { const reason = guard(exec) if (reason !== undefined) return reason } if (exec.agent !== undefined) { for (const { guard } of this.scopedGuards.get(exec.agent) ?? []) { const reason = guard(exec) if (reason !== undefined) return reason } } return undefined } /** Whether every restriction registered for `scope` admits the global tool `name` (intersection semantics). */ private admits(scope: ScopeKey | undefined, name: string): boolean { if (scope === undefined) return true const filters = this.restrictions.get(scope) if (!filters) return true return filters.every(filter => (filter.allow === undefined || filter.allow.has(name)) && (filter.deny === undefined || !filter.deny.has(name))) } /** * Resolve every registry fact one scope needs in one layer traversal. The * visible map applies global restrictions, scoped shadowing, and the reserved * presentation transport; the other sets retain the pre-restriction facts * needed by restriction and prompt-order validation. * @param scope - the viewing scope (the agent), or undefined for the global view. * @returns the complete derived view for that scope. */ private view(scope?: ScopeKey): ToolView { const layer = scope === undefined ? undefined : this.scoped.get(scope) const visible = new Map() const knownNames = new Set() const restrictableNames = new Set() for (const [name, definition] of this.global) { knownNames.add(name) restrictableNames.add(name) if (this.admits(scope, name)) visible.set(name, definition) } // Scoped layer second: same-name entries REPLACE (shadow) the global ones, // and scope-local registrations are never part of the global filter above. for (const [name, definition] of layer ?? []) { knownNames.add(name) visible.set(name, definition) } // Presentation infrastructure is resolved last and outside capability // filtering. Registration rejects this reserved name, so the insertion is // an invariant assertion as well as protection against future layer changes. if (this.codeTransport !== undefined) { visible.set(RUN_CODE_NAME, this.codeTransport) } return { visible, knownNames, restrictableNames } } /** * Look up a tool as one scope sees it (scoped * shadows global; a restricted-away global reads as absent). Presenters pass * the calling agent so the rendered card matches the definition that * actually executed. * @param name - the tool name as registered. * @param scope - the viewing scope (the agent); omitted = the global view. * @returns the definition the scope resolves, or undefined when none is visible. */ get(name: string, scope?: ScopeKey): ToolDefinition | undefined { return this.view(scope).visible.get(name) } /** * The model-facing schemas of everything `scope` can see — exactly the * fields (`name`, `description`, `parameters`) this registry contributes to * system-prompt assembly before its expert transformation waterfall. * Constructed EXPLICITLY rather than by stripping * known non-schema members: a `ToolDefinition` also carries `execute` and the * optional `presentCall`/`presentResult` UI callbacks, and those (especially * the functions) must never leak into a model request. An allowlist can't * drift when a new non-schema member is added to the definition; a denylist * (rest-destructure) would silently leak it. * @param scope - the viewing scope (the agent); omitted = the global view. * @returns one deep-cloned schema per visible tool. */ schemas(scope?: ScopeKey): ToolSchema[] { return [...this.view(scope).visible.values()].map(definition => this.schemaOf(definition, true)) } /** Project one definition onto the model-facing schema fields. */ private schemaOf(definition: ToolDefinition, detachParameters: boolean): ToolSchema { const { name, description, parameters } = definition return { name, description, parameters: detachParameters ? structuredClone(parameters) : parameters, } } /** * Execute one tool call through the `tools/pre-execute` → guards → * `tools/execute` (around dispatch) → `tools/post-execute` → `tools/result` * pipeline. `pre-execute` is the extensible gate * (allow/deny/ask), `tools/execute` wraps core dispatch (a timeout/retry/metrics * seam), and `post-execute` is the inspect/transform seam; core dispatch sits * as the base `next()` of the `tools/execute` waterfall. The whole thing is * wrapped in one outer try/catch so a throwing listener (in any waterfall) * becomes an `isError` result instead of failing the turn; the tool body ALSO * keeps its own inner try/catch, so a thrown tool becomes an `isError` result * that `tools/execute` and `post-execute` listeners can still inspect. If the * tool is not registered (or not visible to the calling agent — a * restricted-away global is exactly as absent as a nonexistent one), the * result is an `isError` carrying a `UNKNOWN_TOOL` structured error. A thrown * {@link HarnessError} surfaces its `{ name, code }` on the result. Before * the final observe-only notification, the authoritative outcome is * materialized as a detached lossless-JSON snapshot; an invalid outcome is * normalized to an error. * @param exec - the typed same-process call input. The registry assigns its * correlation token before policy begins. * @returns the materialized final result after every waterfall; listener and * tool failures resolve as `isError` results rather than rejections. */ async execute(exec: ToolExecutionInput): Promise { const token = createExecutionToken() const callId = exec.callId const name = exec.name const agent = exec.agent const parent = exec.parent const signal = exec.signal const base = { token, callId, name, ...agent !== undefined ? { agent } : {}, ...parent !== undefined ? { parent } : {}, ...signal !== undefined ? { signal } : {}, } let execution: ToolExecution try { const detached = snapshotJsonValue(exec.arguments) if (detached === undefined) { throw new TypeError('tool execution arguments must be losslessly JSON-serializable') } execution = { ...base, arguments: deepFreeze(detached), } } catch (error: unknown) { execution = { ...base, arguments: undefined } const result = this.materializeFinalResult(toolErrorResult(callId, error)) this.notifyResult(execution, result) return result } let result: ToolExecutionResult try { result = this.materializeFinalResult(await this.executePipeline(execution)) } catch (error: unknown) { // Outer backstop: a throwing pre/post-execute listener, guard, or the // waterfall machinery becomes an isError result, never a turn failure. result = this.materializeFinalResult(toolErrorResult(execution.callId, error)) } this.notifyResult(execution, result) return result } /** Run the transformable pipeline; {@link execute} owns final normalization and notification. */ private async executePipeline(exec: ToolExecution): Promise { // --- Gate: tools/pre-execute. An `ask` resolves through the optional // approval seam (or degrades to deny) before the monotonic guards run. The // carrier keys dispatch by exec.agent, so an `agent.ctx` listener gates only // its own agent's calls (agent-less calls are subject-less). const carrier = scopeTarget(this, exec.agent) const gate = await this.ctx.waterfall( carrier, 'tools/pre-execute', exec, () => Promise.resolve({ kind: 'allow' }), ) const decision = gate.kind === 'ask' ? await this.serviceAsk(exec, gate) : gate const denialReason = decision.kind === 'allow' ? this.guardReason(exec) : decision.reason if (denialReason !== undefined) { // Every non-grant, including a failed/unavailable approval request, takes // the same deny path and still reaches post-policy plus result observers. const denied: ToolExecutionResult = { callId: exec.callId, content: [{ type: 'text', text: `Error: ${denialReason}` }], isError: true, } return await this.postExecute(exec, denied) } // --- Around-dispatch: tools/execute. The base `next` is the dispatch- // with-normalization thunk — the tool body's own try/catch turns a throw // into an isError result so a wrapper (and post-execute) can inspect it; // an unknown tool routes through the same catch. A `tools/execute` listener // (e.g. a timeout plugin) wraps this thunk: it may replace `exec.signal` // before delegating and inspect the normalized result after. Dispatched with the // same carrier as the gate, so an `agent.ctx` wrapper wraps only its own // agent's calls. --- const result = await this.ctx.waterfall( carrier, 'tools/execute', exec, async (): Promise => { try { // Resolve through the CALLER's visible view ({@link get}): a scoped // tool shadows its global name-twin for that agent, and a // restricted-away global tool is exactly as absent as a nonexistent // one — same UNKNOWN_TOOL result, no capability leak in the error. const tool = this.get(exec.name, exec.agent) if (!tool) throw new ToolNotFoundError(exec.name) // Normalize the two `execute` return shapes: a bare ContentBlock[] (no // meta) or a { content, meta } object (a tool attaching a private // presentation payload). An array IS the content; the object carries it. const returned = await tool.execute(exec.arguments, exec) const content = Array.isArray(returned) ? returned : returned.content const meta = Array.isArray(returned) ? undefined : returned.meta return { callId: exec.callId, content, isError: false, ...meta !== undefined ? { meta } : {} } } catch (error: unknown) { return toolErrorResult(exec.callId, error) } }, ) if (result.callId !== exec.callId) { throw new TypeError(`tools/execute returned callId "${String(result.callId)}" for authoritative call "${exec.callId}"`) } return await this.postExecute(exec, result) } /** Notify final-result observers without giving them a mutation/error channel into the outcome. */ private notifyResult(exec: ToolExecution, result: ToolExecutionResult): void { // The pipeline is over: freeze the remaining mutable signal slot so every // observer sees the SAME WeakMap-keyable execution without a mutation race. Object.freeze(exec) const callbacks = this.ctx.events.dispatch('emit', [ scopeTarget(this, exec.agent), 'tools/result', exec, result, ]) for (const callback of callbacks) { try { callback(exec, result) } catch (error: unknown) { this.ctx.logger.warn(`tool "${exec.name}" (${exec.callId}): tools/result observer failed: ${errorMessage(error)}`) } } } /** * Resolve an `ask` decision to allow/deny through the approval seam. The * seam is consumed opportunistically with `ctx.get('approval')` — a * deployment that composes no ApprovalService keeps the historical degrade * to deny, and an unmount mid-session degrades the same way on the next ask. * An agent-less execution also degrades: without an agent there is no * session to audit to and no UI to route to. Otherwise the outcome maps * one-to-one — `allowed-once` proceeds; the three non-grants deny with * distinct reasons so the model can tell a human "no" from an absent * approval channel. */ private async serviceAsk( exec: ToolExecution, ask: Extract, ): Promise> { const approval = this.ctx.get('approval') if (approval === undefined) { return { kind: 'deny', reason: ask.reason ?? `tool "${exec.name}" requires approval (not yet supported)` } } if (exec.agent === undefined) { return { kind: 'deny', reason: `tool "${exec.name}" requires approval, but the call has no agent to route it through` } } const outcome = await approval.request({ agent: exec.agent, toolName: exec.name, callId: exec.callId, ...ask.reason !== undefined ? { reason: ask.reason } : {}, ...exec.signal !== undefined ? { signal: exec.signal } : {}, }) switch (outcome) { case 'allowed-once': return { kind: 'allow' } case 'rejected': return { kind: 'deny', reason: `the user rejected tool "${exec.name}"` } case 'cancelled': return { kind: 'deny', reason: `approval for tool "${exec.name}" was cancelled` } case 'unavailable': return { kind: 'deny', reason: `tool "${exec.name}" requires approval, but no approval channel is available` } default: return assertNever(outcome, 'ApprovalOutcome') } } /** * Run the `tools/post-execute` waterfall over a dispatched `result` and apply * its {@link PostToolDecision}: `accept` keeps the call successful (replacing * `content` when given), `block` turns it into an `isError` whose content is * the corrective `feedback`. Either decision may attach `additionalContext`, * which is ferried on the returned result for the loop's per-step buffer. * Runs inside `execute`'s outer try/catch (a throwing listener → isError). */ private async postExecute(exec: ToolExecution, result: ToolExecutionResult): Promise { const decision = await this.ctx.waterfall( scopeTarget(this, exec.agent), 'tools/post-execute', exec, result, () => Promise.resolve({ kind: 'accept' }), ) const additionalContext = decision.additionalContext if (decision.kind === 'block') { return { callId: result.callId, content: decision.feedback, isError: true, ...additionalContext ? { additionalContext } : {}, } } // Accept: replace content if supplied and preserve the dispatched outcome. return { ...result, ...decision.content ? { content: decision.content } : {}, ...additionalContext ? { additionalContext } : {}, } } /** Materialize the authoritative commit outcome once, immediately before `tools/result`. */ private materializeFinalResult(result: ToolExecutionResult): ToolExecutionResult { const detached = snapshotJsonValue(result) if (detached === undefined) { throw new TypeError('tool result must be losslessly JSON-serializable') } return deepFreeze(detached) } } /** Mint a same-process correlation token whose identity is its value. */ function createExecutionToken(): ToolExecutionToken { return Symbol('dsh.tool.execution') as ToolExecutionToken } function toolErrorResult(callId: ToolExecution['callId'], error: unknown): ToolExecutionResult { const info = errorInfo(error) return { callId, content: [{ type: 'text', text: `Error: ${errorMessage(error)}` }], isError: true, ...info ? { error: info } : {}, } } export default ToolRegistry