Merge remote-tracking branch 'origin/feat/send-unify' into xtr/agent-loop-message-machine

# Conflicts:
#	.agents/notes/implemented/architecture/2026-07-22-unified-send-and-coalesced-user-messages.i18n.yaml
#	.agents/notes/implemented/architecture/2026-07-22-unified-send-and-coalesced-user-messages.md
#	.agents/notes/implemented/architecture/2026-07-22-unified-send-and-coalesced-user-messages.zh.md
#	packages/context/workspace-context/README.md
#	packages/core/agent-loop/src/agent.ts
#	packages/core/agent-loop/src/inbox.ts
#	packages/core/agent-loop/tests/agent.spec.ts
#	packages/core/agent-loop/tests/inbox.spec.ts
#	packages/goal/goal/README.md
This commit is contained in:
_Kerman
2026-07-24 14:14:12 +08:00
9 changed files with 54 additions and 46 deletions

View File

@@ -33,7 +33,7 @@ The private per-session cache is keyed by `session.surface.replaceGeneration` an
## Surface contract
`SurfaceEventType` is a closed union — only `user/message`, `assistant/message`, `tool/result`, `context/message`, and `steering/message` may carry `surfaceOp`. A `compact/*` event therefore **cannot** appear on the surface. A successful compaction instead:
`SurfaceEventType` is a closed union — only `user/message`, `assistant/message`, `tool/result`, and `steering/message` may carry `surfaceOp`. A `compact/*` event therefore **cannot** appear on the surface. A successful compaction instead:
1. appends `compact/start` (log-only) — acquires the lock,
2. summarizes the range,

View File

@@ -18,9 +18,9 @@ When `timeZone` is omitted, the plugin resolves the Node process's system zone o
## Timing semantics
The plugin prepends an `agent/pre-step` listener. When an injection is due, it appends one `context/message` through `agent.inject()` before `step/start` and ordinary automatic compaction, with source `{ kind: 'plugin', plugin: 'time-context' }`. A suppressed attempt appends nothing.
The plugin prepends an `agent/pre-step` listener. When an injection is due, it appends one injected `user/message` through `agent.inject()` before `step/start` and ordinary automatic compaction, with source `{ kind: 'plugin', plugin: 'time-context' }`. A suppressed attempt appends nothing.
Positive-interval scheduling scans the raw durable session events for the latest `context/message` with that source, including a reading shadowed by compaction. The schedule therefore applies across turns and resumed processes without process-local cache state. It reduces append frequency and history growth but never removes an existing reading, and sessions schedule independently.
Positive-interval scheduling scans the raw durable session events for the latest `user/message` with that source, including a reading shadowed by compaction. The schedule therefore applies across turns and resumed processes without process-local cache state. It reduces append frequency and history growth but never removes an existing reading, and sessions schedule independently.
Step 1 measures from the latest preceding model-visible message, including the prompt that opened the turn. Later steps measure from the preceding time-context event in the same turn. Both baselines use durable session-event timestamps; backward wall-clock movement clamps elapsed time to zero. A missing first-step baseline, or a later step with no earlier same-turn reading because interval suppression skipped it, reports `unavailable`.

View File

@@ -899,8 +899,8 @@ export const EVENT_API: readonly EventApiEntry[] = [
name: 'agent/inbox/discard',
mode: 'emit',
signature: '\'agent/inbox/discard\'(this: Scoped<Agent>, agent: Agent, messages: AgentMessage[]): void',
jsDoc: '/**\n * `cancel()` (without `keepInbox`) dropped pending inbox items without\n * delivering them. Fires once per effective clearing call with every\n * discarded item, after `agent/cancel-requested` and before the abort.\n * @param agent - the agent whose inbox was cleared.\n * @param messages - the discarded messages in FIFO order (queued then steering); empty when nothing was pending.\n * Scope-filtered dispatch (`@deepseek-ai/dsh-scope`): agent-scoped listeners receive only that agent.\n * @mode emit\n */',
summary: '`cancel()` (without `keepInbox`) dropped pending inbox items without delivering them.',
jsDoc: '/**\n * Pending inbox items were dropped without delivering them, so every\n * enqueued id receives exactly one terminal `agent/inbox/dequeue` OR\n * `agent/inbox/discard`. Emitters: `cancel()` without `keepInbox` (after\n * `agent/cancel-requested`, before the abort); a terminal `agent/turn-stop`\n * dropping pending steering (in-turn and on the post-turn late-steering\n * drain); and disposal of any still-pending items (before\n * `agent/status(\'disposed\')`). Fires once per drop with every dropped item.\n * @param agent - the agent whose inbox items were dropped.\n * @param messages - the discarded messages in FIFO order (queued then steering); never empty.\n * Scope-filtered dispatch (`@deepseek-ai/dsh-scope`): agent-scoped listeners receive only that agent.\n * @mode emit\n */',
summary: 'Pending inbox items were dropped without delivering them, so every enqueued id receives exactly one terminal `agent/inbox/dequeue` OR `agent/inbox/discard`.',
},
{
name: 'agent/inbox/enqueue',

View File

@@ -338,11 +338,15 @@ declare module 'cordis' {
*/
'agent/inbox/dequeue'(this: Scoped<Agent>, agent: Agent, message: AgentMessage): void
/**
* `cancel()` (without `keepInbox`) dropped pending inbox items without
* delivering them. Fires once per effective clearing call with every
* discarded item, after `agent/cancel-requested` and before the abort.
* @param agent - the agent whose inbox was cleared.
* @param messages - the discarded messages in FIFO order (queued then steering); empty when nothing was pending.
* Pending inbox items were dropped without delivering them, so every
* enqueued id receives exactly one terminal `agent/inbox/dequeue` OR
* `agent/inbox/discard`. Emitters: `cancel()` without `keepInbox` (after
* `agent/cancel-requested`, before the abort); a terminal `agent/turn-stop`
* dropping pending steering (in-turn and on the post-turn late-steering
* drain); and disposal of any still-pending items (before
* `agent/status('disposed')`). Fires once per drop with every dropped item.
* @param agent - the agent whose inbox items were dropped.
* @param messages - the discarded messages in FIFO order (queued then steering); never empty.
* Scope-filtered dispatch (`@deepseek-ai/dsh-scope`): agent-scoped listeners receive only that agent.
* @mode emit
*/

View File

@@ -30,11 +30,11 @@ The chain key is `(tool name, canonical arguments)` — canonicalization is a de
## Reminder delivery
Reminders ride the post-execute decision's `additionalContexts` (source `{kind: 'plugin', plugin: 'repeat-tool-guard'}`), never a `content` replacement: the `tool/result` event stays the tool's own output for audit. The loop buffers the context and appends it as a `context/message` after the step's tool results, which the session renders as a plain synthetic user message — so the reminder is model-visible, source-attributed, and reconstructable from the session log with no new session event. The guard always delegates via `next()` and prepends its reminder to the downstream decision's context array (both variants — a blocked call still gets the nudge); every entry retains its own source and metadata.
Reminders ride the post-execute decision's `additionalContexts` (source `{kind: 'plugin', plugin: 'repeat-tool-guard'}`), never a `content` replacement: the `tool/result` event stays the tool's own output for audit. The loop buffers the context and appends it as an injected `user/message` after the step's tool results, which the session renders as a plain synthetic user message — so the reminder is model-visible, source-attributed, and reconstructable from the session log with no new session event. The guard always delegates via `next()` and prepends its reminder to the downstream decision's context array (both variants — a blocked call still gets the nudge); every entry retains its own source and metadata.
## Testing
Unit suites drive a real agent loop against a mock adapter (no network) and cover the chain semantics above to per-file 100%. The snapshot tier owns the transcript surface: a scripted-replay scenario repeats a call five times and pins both reminder tiers (gentle at 3, detailed at 5) as `context/message`s in the ACP transcript.
Unit suites drive a real agent loop against a mock adapter (no network) and cover the chain semantics above to per-file 100%. The snapshot tier owns the transcript surface: a scripted-replay scenario repeats a call five times and pins both reminder tiers (gentle at 3, detailed at 5) as injected `user/message`s in the ACP transcript.
## Model Experience

View File

@@ -27,7 +27,7 @@ Why a shared lib at all: Codex deliberately reimplements a *subset* of the Claud
Declaration-merged into `SessionEventMap` (log-only, like `compact/*` — NOT a `SurfaceEventType`, no `surfaceOp`): `hook/invoked` (a hook command ran) and `hook/result` (its outcome, paired by `handlerId`, with `appendHookResult` owning the decision rule). Payloads and per-event JSDoc are in the generated [persistence log event catalog](../../../docs/persistence-catalog.md); `stderrSummary` is truncated to the record's `stderrSummaryMaxChars` (the bridge's config, reference default `DEFAULT_STDERR_SUMMARY_MAX_CHARS` = 500; omitted when empty).
Like every event they must sit inside an open turn. The mid-turn points (`PreToolUse`/`PostToolUse`/`UserPromptSubmit`/`Stop`) fire inside the loop's open turn by construction; `SessionStart` gets no `hook/*` record (its injected `context/message` is the durable evidence) — see the hooks Agent Note.
Like every event they must sit inside an open turn. The mid-turn points (`PreToolUse`/`PostToolUse`/`UserPromptSubmit`/`Stop`) fire inside the loop's open turn by construction; `SessionStart` gets no `hook/*` record (its injected `user/message` is the durable evidence) — see the hooks Agent Note.
## Model Experience

View File

@@ -6,7 +6,7 @@ Logged, per-agent plan collaboration state with deployment-owned guidance, direc
`plan/mode` (`{ active: boolean }`) is a log-only, whole-value-replace `SessionEventMap` member. `foldPlanMode(events)` returns the last logged value or `false`, so resume, fork, and compaction recover plan state directly from the session log. UIs observe committed flips through `session/event`.
`ctx.planMode.set(agent, active)` records a pending selection and flushes it inside the next turn boundary. `get(agent)` returns `{ active, pending? }`, separating the logged state shaping the current step from a user's optimistic selection. Prompt submission, ordinary continuation, and request-recovery retry are all covered; a changed user selection contributes one `context/message` notice when the last logged request header described the other state.
`ctx.planMode.set(agent, active)` records a pending selection and flushes it inside the next turn boundary. `get(agent)` returns `{ active, pending? }`, separating the logged state shaping the current step from a user's optimistic selection. Prompt submission, ordinary continuation, and request-recovery retry are all covered; a changed user selection contributes one plugin-sourced `user/message` notice when the last logged request header described the other state.
## Model and human surfaces