Files

Tianyi Cui 9389c5df19 docs(rfc): address Codex review findings on the RFC batch

Two overstated evidence claims, reworded to the precise fact:
- prune-dead-core-spine-surface: runLoop has no importer OUTSIDE the
  package (its own agent.ts drives the loop with it); the dead surface
  is the public re-export, not the function.
- generic-long-running-tool-runtime census: BashTask.done has no
  consumer through the public seam, but is production-load-bearing
  inside dsh-bash-local (disposal awaits it); only the public
  completion surface is single-consumer.

Also fold the reviewer-located doc sites into the removal sets so the
implementing PRs need no re-discovery: the llm/pi-ai/compact-basic
README rows and the adding-an-llm-adapter cookbook line (prefill/image),
the content-block-vocabulary RFC's has-a-home consequence lines
(cache/prefill/image), the tools.md paste + type-equiv manifest row +
tools README row (callId), and the session-surface RFC's
full-rebuild-after-replacement sentence (invalidate).

2026-07-04 04:21:38 +08:00

4.5 KiB

Raw Blame History

RFC: Extract a generic long-running tool runtime

Status: proposed

Problem

The bash capability seam supports both foreground commands and long-running background tasks. Background support is large: the abstract executor exposes start, get, ownerOf, list, readOutput, kill, and onTaskDone; the local executor tracks tasks, incremental reads, owner tokens, process cleanup, and completion listeners; the model sees three tools (bash, bash_output, bash_kill); the tool plugin injects completion notices back into the owning agent's session. The local executor fences task access behind owner tokens because predictable global task ids are a cross-session read/kill hazard.

The tool cookbook already points at the real design smell: background bash is really generic long-running-tool infrastructure living inside one tool. If future tools need background execution, polling, kill, ownership, and completion notices, those semantics should not be hidden in dsh-bash.

Proposal

Move long-running task semantics above bash into a tool-agnostic runtime. Bash remains able to run background commands, but it stops owning the general concepts of task ids, ownership tokens, polling, cancellation, completion notifications, and model-facing "read/kill this task" commands.

The runtime should own:

Stable task ids and owner tokens keyed to the calling session/agent.
Registration of a long-running task with a producer for incremental output and a completion promise.
Generic read/cancel/list operations with the same cross-session authorization rule for every tool.
Completion notification injection into the owning session.
Presentation hooks for pending/running/completed task state, with bash supplying only command-specific labels and output formatting.

dsh-bash then keeps the bash-specific execution contract: resolve a request into a command spec, run a foreground command, or start a process and hand its streams/process handle to the generic runtime. dsh-tool-bash keeps the model-facing command tool, but the follow-up operations become generic long-running-tool operations or a shared utility that bash registers with, rather than bespoke bash_output/bash_kill plumbing.

Current seam consumption

A consumer census of the surface the runtime would carve up. Production (packages/bash/tool-bash/src/index.ts) consumes resolve, run, start, ownerOf, readOutput, kill, and onTaskDone. get()/list() have test-harness consumers only — they were removed once and reverted on the merits (the implementation note in prune dead methods from the persistence seam records the test-migration cost dwarfing the surface removed). The per-task BashTask.done promise has no consumer through the public seam either (dsh-tool-bash completes via onTaskDone), but it is production-load-bearing INSIDE the implementation: dsh-bash-local's disposal awaits it to reach quiescence. The seam therefore exposes two public completion representations — the per-task promise and the global onTaskDone listener registry — and the shipped consumer uses only the latter: the runtime should pick exactly one public completion surface and record which. One shape wart for the split to dissolve: BashExecSpec.timeoutMs is required but ignored by start(), an artifact of sharing one spec type between foreground and background execution.

Acceptance criteria

The bash-specific packages no longer define the generic task registry, owner-token authorization, polling, cancellation, or completion-notification machinery.
A shared long-running-task service or tool layer owns those semantics and is documented as the path for any future background-capable tool.
Bash background behavior remains available through the shared layer, with tests proving cross-session isolation still holds.
ACP and snapshot fixtures render background bash through the shared task vocabulary, not through bash-only lifecycle semantics.
The tool cookbook points long-running tools at the shared runtime instead of telling each tool to invent its own task protocol.

What we give up

The bash package loses local ownership of an already-working background-task implementation, and the implementing PR may temporarily churn model-facing tool names or transcript presentation. That churn is worthwhile if it leaves one background-task contract instead of making every future long-running tool clone bash's private protocol.

4.5 KiB Raw Blame History