feat(subagent): catalog one-shot child sessions

This commit is contained in:
Dudu-0223
2026-07-27 16:58:37 +08:00
committed by Tianyi Cui
parent de6e572e51
commit 774ee34b9a
85 changed files with 1034 additions and 539 deletions

View File

@@ -28,13 +28,13 @@ Multiple providers may coexist under different names. This lets a deployment exp
| `registerProvider(provider)` | Register one trusted same-process implementation by name. Registration is effect-scoped; removing it prevents new starts but does not revoke runs already returned to callers. Duplicate names fail loud. |
| `getProvider(name)` | Return the provider, or `undefined` when absent. |
| `list()` | Return provider names in insertion order. |
| `start(name, request)` | Validate an ordinary caller request, then await the provider until a real one-shot child is ready. Fulfillment returns a holder-owned `SubagentRun`; rejection means the provider has already cleaned every partial startup resource. Continuable children never enter through this operation. |
| `start(name, request)` | Validate an ordinary caller request, resolve its detached `one-shot` descriptor, then await the provider until a real one-shot child is ready. Fulfillment returns a holder-owned `SubagentRun`; rejection means the provider has already cleaned every partial startup resource. Continuable children never enter through this operation. |
| `startContinuable(spec)` | Establish one durable continuable child and deliver its initial prompt. Resolves with `{ childId, messageId }` when the child's inbox accepts that prompt, without waiting for the turn to start or for the message to reach the Session log; any earlier failure rejects with no ids and rolls the child back entirely. Requires `ctx.agents`, session persistence, and a provider with the `prepareContinuable` capability. |
| `followup(parent, childId, content, { source, signal })` | Deliver one later message from the exact live direct parent as the child's next FIFO turn, matching `Agent.followup()` terminology, and return the accepted `MessageId`. A resident child's inbox accepts it directly (waking a waiting Activation); an absent one cold-resumes from its persisted Session. Requires `ctx.agents`; cold resume also requires session persistence. |
| `drainContinuableDescendants(parents)` | Close admission below exact live host-owned parent Agents, stop only their visible continuable descendants, await materializations admitted below those roots through publication or rollback, then release the selected forests child-first. The cutoff lasts until each exact parent leaves the registry; unrelated parent forests and manager-wide admission remain live. |
| `listChildren(parentSessionId, signal?)` | List direct continuable children and per-child diagnostics in stable trace order without loading or resuming them. Requires session query; it does not require `ctx.agents` or the continuation manager. |
| `listChildren(parentSessionId, signal?)` | List direct session-backed subagents with their `one-shot`/`continuable` mode and `running`/`inactive` activity, plus per-child diagnostics, in stable trace order without loading or resuming them. Requires session query; it does not require `ctx.agents` or the continuation manager. |
`SubagentStartRequest.signal` is required and is the canonical cancellation channel for a one-shot `start`. An abort before publication makes `start()` reject after rollback; an abort after publication cancels the live child. The request may also select a model, require structured output, cap delegation depth, restrict child tools, or set a child persona. For a continuable start or follow-up, the caller signal owns lookup, materialization, and admission only until inbox acceptance; afterward the manager owns the Activation independently, so later caller cancellation neither cancels the accepted turn nor disposes the child.
`SubagentStartRequest.label` is the short durable display label for a session-backed child. `signal` is required and is the canonical cancellation channel for a one-shot `start`. An abort before publication makes `start()` reject after rollback; an abort after publication cancels the live child. The request may also select a model, require structured output, cap delegation depth, restrict child tools, or set a child persona. For a continuable start or follow-up, the caller signal owns lookup, materialization, and admission only until inbox acceptance; afterward the manager owns the Activation independently, so later caller cancellation neither cancels the accepted turn nor disposes the child.
Follow-up authority comes from the exact live direct parent recorded in the child's durable header. Cold resume checks that authority before reconstruction and again in the final no-await inbox-admission span, so a parent unregistered or replaced during materialization cannot authorize delivery. The `source` on a follow-up is durable provenance retained on the delivered message and grants no authority.
@@ -53,7 +53,7 @@ Continuable creation is the optional `SubagentProvider.prepareContinuable?()` me
## The durable descriptor
The seam owns the versioned `subagent/descriptor` session event vocabulary (`src/descriptor.ts`): `snapshotSubagentDescriptor()` validates and detaches the declared composition before the child session exists, and `foldSubagentDescriptor()` validates the complete current-version payload before recovering it from a loaded child log. Malformed current-version payloads fail before materialization; unsupported versions make the child non-resumable. The payload records the provider name, resolved child `agentOptions.provider`/`model`, and optional `persona`/`toolFilter` — explicit fields, never the merge-extensible `AgentOptions` object, so an unrelated extension value cannot break continuation. It omits `subagentDepth` (the persisted header's `delegationDepth` is the monotone floor) and `outputSchema` (never captured for a continuable child). The event is log-only: no `surfaceOp`, absent from model history, and retained by the append-only log across compaction.
The seam owns the versioned `subagent/descriptor` session event vocabulary (`src/descriptor.ts`): `snapshotSubagentDescriptor()` validates and detaches the record before provider work, and `foldSubagentDescriptor()` validates the complete current-version payload before recovering it from a loaded child log. Every local session-backed start appends one descriptor with the durable creation label, provider name, and lifecycle `mode`. A `one-shot` descriptor stops there; a `continuable` descriptor additionally records resolved child `agentOptions.provider`/`model` and optional `persona`/`toolFilter` for cold resume. These are explicit fields, never the merge-extensible `AgentOptions` object, so an unrelated extension value cannot break continuation. The descriptor omits `subagentDepth` (the persisted header's `delegationDepth` is the monotone floor) and `outputSchema` (an activation's result contract). The event is log-only: no `surfaceOp`, absent from model history, and retained by the append-only log across compaction. Malformed current-version payloads are corrupt; unsupported versions cannot be classified by this runtime.
## Delegation depth
@@ -67,7 +67,7 @@ The seam owns the depth vocabulary shared by implementations and consumers: the
`SubagentRun.result` resolves to `{ output, structured?, stopReason }`. Child-level failures resolve with a non-`completed` reason; only an infrastructure fault that the seam cannot represent may reject. `dispose()` is idempotent, cancels remaining work, and waits for the child resources to quiesce.
A local run publishes an ordinary child agent/session before `start()` fulfills, returns that shared session id as `SubagentRun.id`, exposes the exact child as `SubagentRun.localAgent`, and records `request.parent.session.id` in the child's `parentSession` header. Remote providers instead mint a parent-scoped lifecycle id and return `localAgent: undefined`.
A local run publishes an ordinary child agent/session before `start()` fulfills, returns that shared session id as `SubagentRun.id`, exposes the exact child as `SubagentRun.localAgent`, records `request.parent.session.id` in the child's `parentSession` header, and appends the resolved descriptor inside its initial turn. Remote providers instead mint a parent-scoped lifecycle id and return `localAgent: undefined`; without a local child session, their one-shot runs are not part of trace-backed enumeration.
## Continuable children and Activations
@@ -89,7 +89,9 @@ Provider additions and removals also emit `subagent/provider-added` and `subagen
## Collection model
The model-facing tool collects synchronously by default: it awaits the child result and disposes the run before returning. One-shot background delegation registers a plain Task in the tool, whose generic status, collection, and cancellation tools own later interaction. Continuable background delegation calls `ctx.subagents.startContinuable()` and returns only the durable child id; the child owns its own turns from inbox acceptance, so there is no Task, no result promise, and no public subagent cancellation — a caller sends later work with the `send_message` follow-up tool, and the durable child Session remains the source of the child's detailed output. The continuation manager exists only while `ctx.agents` is available, and session persistence is resolved per continuation operation. Independently, `listChildren()` resolves session query and dynamically imports its optional runtime only when called, then interprets a read-only live-preferred scan without consulting the continuation manager, Agent registrations, Activations, or providers. It forwards the caller's signal to cancellable trace and exact-read operations, checks cancellation around the remaining event-list read, and reports every observed abort as `SubagentError` code `CANCELLED`. See the [background subagent tasks Agent Note](../../../.agents/notes/implemented/feature/2026-07-08-background-subagent-tasks.md), the [continuable background subagents Agent Note](../../../.agents/notes/implemented/feature/2026-07-21-continuable-background-subagents.md), the [durable catalog Agent Note](../../../.agents/notes/implemented/feature/2026-07-22-durable-subagent-catalog-and-list-agents.md), the [merged-service Agent Note](../../../.agents/notes/implemented/simplification/2026-07-26-merge-subagent-control-service.md), the [capability-seam Agent Note](../../../.agents/notes/implemented/feature/2026-06-21-subagent-capability-seam.md), and `src/types.ts` for the complete contracts.
The model-facing tool collects synchronously by default: it awaits the child result and disposes the run before returning. One-shot background delegation registers a plain Task in the tool, whose generic status, collection, and cancellation tools own later interaction. Continuable background delegation calls `ctx.subagents.startContinuable()` and returns only the durable child id; the child owns its own turns from inbox acceptance, so there is no Task, no result promise, and no public subagent cancellation — a caller sends later work with the `send_message` follow-up tool, and the durable child Session remains the source of the child's detailed output. The continuation manager exists only while `ctx.agents` is available, and session persistence is resolved per continuation operation. Independently, `listChildren()` resolves session query and dynamically imports its optional runtime only when called, then interprets a read-only live-preferred scan of all descriptor-bearing direct children without consulting the continuation manager, Agent registrations, Activations, or providers. Service consumers such as a UI can retain both modes; the model-facing `list_agents` tool projects only `continuable` entries and maps service activity to its existing `running`/`complete` vocabulary. The scan forwards the caller's signal to cancellable trace and exact-read operations, checks cancellation around the remaining event-list read, and reports every observed abort as `SubagentError` code `CANCELLED`. See the [background subagent tasks Agent Note](../../../.agents/notes/implemented/feature/2026-07-08-background-subagent-tasks.md), the [continuable background subagents Agent Note](../../../.agents/notes/implemented/feature/2026-07-21-continuable-background-subagents.md), the [durable catalog Agent Note](../../../.agents/notes/implemented/feature/2026-07-22-durable-subagent-catalog-and-list-agents.md), the [merged-service Agent Note](../../../.agents/notes/implemented/simplification/2026-07-26-merge-subagent-control-service.md), the [capability-seam Agent Note](../../../.agents/notes/implemented/feature/2026-06-21-subagent-capability-seam.md), and `src/types.ts` for the complete contracts.
Continuable Activations require final durability confirmation. One-shot runs retain best-effort session checkpointing, so a completed one-shot child is discoverable after disposal only when its session actually reached persistence; the service does not invent a catalog entry from Task history when that checkpoint is absent.
## Model Experience
@@ -101,9 +103,10 @@ No direct invalidation; the named consumers own any request-prefix changes.
## Known Limitations and Deferred Work
- **ACP children remain one-shot** — an ACP `prepareContinuable` requires persisting the remote session id in provider-specific descriptor data and a per-child continuation advertisement, since ACP `loadSession` support is negotiated per child rather than established by the method's presence. Remote providers also require a separate Activation ownership contract with equivalent authenticated control and child-first quiescence before they support continuable children.
- **ACP children remain one-shot and are not trace-enumerable** — an ACP run has no local child session in the parent's session corpus. An ACP `prepareContinuable` requires persisting the remote session id in provider-specific descriptor data and a per-child continuation advertisement, since ACP `loadSession` support is negotiated per child rather than established by the method's presence. Remote providers also require a separate Activation ownership contract with equivalent authenticated control and child-first quiescence before they support continuable children.
- **No report delivery** — the MVP exposes no `report` tool, child-to-parent content delivery, or automatic parent wakeup; a completed child turn leaves its output in the durable child Session until a caller inspects that transcript or submits another authorized turn.
- **No host-user continuation** — `followup()` requires the exact live direct parent. A future host adapter needs a concrete authenticated interaction before the seam gains a separate user capability.
- **No subagent steering** — every continuation message opens a later FIFO turn, so a parent cannot redirect a turn already underway; the manager stores no current-turn controller state.
- **Process-local residency** — the Activation inbox and ownership graph do not coordinate two harness processes; concurrent access to one persistence store still requires a durable mailbox and cross-process lease protocol.
- **No replay of accepted-but-unlogged messages** — only messages written to the child Session log are reconstructable with their admitted provenance. A crash may lose an accepted initial prompt or follow-up that never reached the log; a later authorized message can cold-resume the child, but the lost message is not replayed automatically.
- **Lifecycle events are observe-only** — a run-affecting `subagent/end` continuation or decision surface waits for a concrete consumer.