feat(subagent): enrich subagent/start + subagent/end lifecycle events (observe-only)

A hooks bridge translating SubagentStart/SubagentStop needs to know WHICH kind of
subagent ran and WHAT it produced — Claude Code's hooks carry subagent_type and the
child's final message. Enrich the existing lifecycle emits to match, observe-only:

- agentType: an optional caller-supplied subagent-kind label (CC's subagent_type),
  added to SubagentStartRequest and carried VERBATIM onto both subagent/start
  (SubagentRunInfo) and subagent/end (SubagentRunEndInfo). The seam never interprets
  it. dsh-tool-subagent threads it from a new optional Config.agentType, so a
  deployment exposing multiple subagent kinds (one tool load per kind) labels each.
- lastAssistantMessage: the child's final output (SubagentResult.output), added to
  SubagentRunEndInfo on the settle path so an observer sees what the subagent
  produced without holding the run. Absent on the reject path (no result produced).

Strictly observe-only: both events stay plain emits (subagent/end fires from a
detached .then and awaits no listener). A control-flow subagent/end (awaited
waterfall returning a decision) would need the emit→waterfall reshape, awaiting
listeners before settling, and a provider resume capability — deferred to the
background/steering redesign (FIXME(subagent-continuation) anchors it). RFC:
implemented/feature/2026-06-30-subagent-observe-enrich.md.
This commit is contained in:
Tianyi Cui
2026-06-30 21:29:08 +08:00
parent 60418a5779
commit 7cc7b9cf7f
11 changed files with 225 additions and 10 deletions

View File

@@ -241,7 +241,7 @@ A subagent run settled — emitted when SubagentRun.result resolves (any stop re
'subagent/end'(info: SubagentRunEndInfo): void
```
Source: [`packages/subagent/subagent/src/index.ts:65`](../../packages/subagent/subagent/src/index.ts)
Source: [`packages/subagent/subagent/src/index.ts:75`](../../packages/subagent/subagent/src/index.ts)
#### `subagent/start` — emit
@@ -251,7 +251,7 @@ A subagent run started — emitted after the provider is resolved and its capabi
'subagent/start'(info: SubagentRunInfo): void
```
Source: [`packages/subagent/subagent/src/index.ts:59`](../../packages/subagent/subagent/src/index.ts)
Source: [`packages/subagent/subagent/src/index.ts:69`](../../packages/subagent/subagent/src/index.ts)
### `system-prompt/*`
@@ -455,7 +455,7 @@ list(): string[]
start(name: string, request: SubagentStartRequest): SubagentRun
```
Source: [`packages/subagent/subagent/src/index.ts:103`](../../packages/subagent/subagent/src/index.ts)
Source: [`packages/subagent/subagent/src/index.ts:130`](../../packages/subagent/subagent/src/index.ts)
### `ctx.systemPrompt` — `SystemPrompt`

View File

@@ -25,6 +25,7 @@ What a caller asks for when starting a subagent. The tool layer builds this from
```ts type-equiv
interface SubagentStartRequest {
prompt: ContentBlock[]
agentType?: string
parent: Agent
signal?: AbortSignal
agentOptions?: AgentOptions
@@ -85,7 +86,7 @@ interface SubagentProvider {
}
```
The service (`ctx.subagents`) emits `subagent/start` when a run begins and `subagent/end` when it settles (see the [events catalog](../cordis-catalog/events-and-services.md)). Both emits contain a thrown listener **per listener** (logged, never propagated): one bad subscriber can neither strand a live run, surface as an unhandled rejection on the detached settle hook, nor starve the listeners registered after it.
The service (`ctx.subagents`) emits `subagent/start` when a run begins and `subagent/end` when it settles (see the [events catalog](../cordis-catalog/events-and-services.md)). Both payloads carry the caller's optional `agentType` label (verbatim from the request — Claude Code's `subagent_type`); `subagent/end` additionally carries `lastAssistantMessage` (the child's final `output`) on the settle path, so an observer sees WHAT the subagent produced without holding the run (absent when the run rejected at the infrastructure level — no result was produced). These are **observe-only** enrichments: both events are plain `emit`s (the `subagent/end` fires from a detached `.then` after the result settles and awaits no listener), so a subscriber observes but cannot change the run. Both emits contain a thrown listener **per listener** (logged, never propagated): one bad subscriber can neither strand a live run, surface as an unhandled rejection on the detached settle hook, nor starve the listeners registered after it.
## In-process backends: depth and seed

View File

@@ -88,6 +88,7 @@ Do NOT write one for a mechanical or local choice (a variable name, a one-file r
| [ACP subagent backend (out-of-process delegation)](implemented/feature/2026-06-22-acp-subagent-backend.md) | 2026-06-22 |
| [The `todo_write` tool — model task list as event-sourced session state](implemented/feature/2026-06-29-todo-write-tool.md) | 2026-06-29 |
| [Interception seams — the typed-Decision surface a hook programs against](implemented/feature/2026-06-30-interception-seams.md) | 2026-06-30 |
| [Subagent lifecycle enrichment — agentType + lastAssistantMessage (observe-only)](implemented/feature/2026-06-30-subagent-observe-enrich.md) | 2026-06-30 |
### Simplification

View File

@@ -0,0 +1,29 @@
# RFC: Subagent lifecycle enrichment — agentType + lastAssistantMessage (observe-only)
Status: implemented (accepted 2026-06-30)
<!-- XXX: legacy ADR/RFC body format, not yet normalized to a unified RFC template. -->
## Context
The hooks subsystem ([interception seams RFC](2026-06-30-interception-seams.md)) lets a plugin observe and gate the agent at lifecycle points. Claude Code and Codex both expose **SubagentStart / SubagentStop** hooks, and CC's carry a `subagent_type` (which named subagent kind ran) and the subagent's final message. The harness already emits `subagent/start` and `subagent/end` lifecycle events ([the subagent capability-seam](2026-06-21-subagent-capability-seam.md)), but their payloads were minimal (`provider`, `id`, and on end `stopReason`) — not enough for a hooks bridge to report which KIND of subagent ran, or WHAT it produced, without separately reaching for the live run.
This RFC enriches those two payloads. It is deliberately **observe-only**: no control-flow change, no waterfall, no `start()` restructure. A run-affecting subagent-stop decision (continuation, injection that changes the run) is a separate, larger redesign and stays out of scope.
## Decision
Add two pieces of information to the subagent lifecycle surface:
1. **`agentType` — a caller-supplied subagent-kind label**, the harness analogue of CC's `subagent_type`. It is optional on `SubagentStartRequest`, carried VERBATIM onto both `subagent/start` (`SubagentRunInfo`) and `subagent/end` (`SubagentRunEndInfo`). The seam never interprets it. The model-facing `dsh-tool-subagent` tool threads it from a new optional `Config.agentType`, so a deployment that exposes multiple subagent kinds (one tool load per kind) labels each. Absent when the caller does not distinguish kinds (the spread omits the key — `exactOptionalPropertyTypes`-correct).
2. **`lastAssistantMessage` — the child's final output**, added to `SubagentRunEndInfo`. On the settle path it is `SubagentResult.output` (so an observer sees WHAT the subagent produced without holding the run). On the REJECT path (an infrastructure fault where no `SubagentResult` was produced — the seam only knows `stopReason: 'error'`) it is absent.
Both events stay plain **`emit`s**. `subagent/end` fires from a detached `.then` on `run.result` and awaits no listener, so it is genuinely observe-only by construction — a `subagent/start` listener can still reach the live child via `ctx.agents.get(info.id)` and `inject()` into it; a `subagent/end` listener can only observe (the run has settled). Per-listener containment (already in place) keeps one bad subscriber from stranding a live run or surfacing as an unhandled rejection on the detached settle hook.
## Why observe-only, and what is deferred
A control-flow `subagent/end` (an awaited waterfall returning a stop/continue decision, like the other interception seams) would require: reshaping `subagent/end` from emit to waterfall, restructuring `SubagentService.start` to await listeners before settling, and implementing the `resume` capability in the in-process provider so a "continue" can actually re-run the child. That belongs to the background/steering subagent redesign the [capability-seam RFC](2026-06-21-subagent-capability-seam.md) already defers (the same redesign that unifies long-running-tool handling across subagents and bash). This RFC ships the observe-only enrichment a hooks bridge needs today; `FIXME(subagent-continuation)` / `TODO` anchors mark where the control-flow version would land if and when that redesign happens.
## Consequences
A hooks bridge (or a native plugin) can now translate SubagentStart/SubagentStop faithfully: it reports `agentType`, matches its hook config on it, and forwards the child's `lastAssistantMessage` to a SubagentStop handler — all by subscribing to the existing emits, no new control-flow surface. The vocabulary addition is documented in [docs/core-data-structures/subagent.md](../../../core-data-structures/subagent.md) (the `SubagentStartRequest` type-equiv block + the events prose) and the two subagent READMEs; the catalog is regenerated. No production behavior changes — the events fire exactly as before, with two more (optional) fields on their payloads — so no snapshot or e2e change is needed.