fix(#4079): forbid ScheduleWakeup wake-tool escape hatch at background-wait sites (#4299)

* test(#4079): guard every background-wait site against ScheduleWakeup wake-tool escape hatch

Content-contract tests over plan-phase.md, stall-detection-helpers.md, and
manager.md: every wait-instruction site must forbid ScheduleWakeup (the
host /loop tool whose partial-args call surfaces the red validation error
in #4079), while the sanctioned wait mechanisms stay byte-preserved.

* fix(#4079): forbid ScheduleWakeup wake-tool escape hatch at every background-wait site

The orchestrator could literalize 'I'll wait' by calling the host's
ScheduleWakeup tool (/loop pacing surface) with partial arguments while a
background subagent was in flight, surfacing the red validation error
'`prompt` is required when `stop` is not true.' GSD never sanctioned a
wake call; now every wait-instruction site says so explicitly: the two
synchronous stop-and-wait ORCHESTRATOR RULEs in plan-phase.md, the
stall-detection-helpers.md fragment (loaded before every gsd_stall_watch
wait), and both manager.md background-dispatch rules. The sanctioned wait
mechanisms (blocking Agent() return, gsd_stall_watch polling, dashboard
loop) are unchanged. The regression test uses the shared
readFileNormalized helper (code-review finding).

Emitted-Drift-Ack-Growth: plan-phase.md — #4079 guard sentence at the two stop-and-wait rules (+396 B, within the phase6 shrink-only line)
Emitted-Drift-Ack-Growth: manager.md — #4079 guard sentence at both background-dispatch rules

* chore(#4079): add changeset fragment (pr placeholder, backfilled after PR open)

* chore(#4079): backfill PR number in changeset

---------

Co-authored-by: sim <sim@local>
This commit is contained in:
Tom Boucher
2026-09-04 23:16:40 -04:00
committed by GitHub
parent 02ad0b91f3
commit c27da5e993
5 changed files with 169 additions and 4 deletions

View File

@@ -274,7 +274,7 @@ Important: You are running in the background. Do NOT use AskUserQuestion — mak
)
```
> **ORCHESTRATOR RULE — BACKGROUND DISPATCH**: After calling Agent() above with `run_in_background=true`, do NOT do any planning work for this phase independently. Return to the dashboard immediately and wait for the background agent to report back. Only resume planning-related work when the subagent result is available.
> **ORCHESTRATOR RULE — BACKGROUND DISPATCH**: After calling Agent() above with `run_in_background=true`, do NOT do any planning work for this phase independently. Return to the dashboard immediately and wait for the background agent to report back. Only resume planning-related work when the subagent result is available. Never call `ScheduleWakeup` or any host wake/sleep-scheduling tool while waiting (#4079) — a partial-args wake call surfaces a red validation error; the dashboard loop is the wait.
Display:
@@ -328,7 +328,7 @@ Important: You are running in the background. Do NOT use AskUserQuestion — mak
)
```
> **ORCHESTRATOR RULE — BACKGROUND DISPATCH**: After calling Agent() above with `run_in_background=true`, do NOT do any execution work for this phase independently. Return to the dashboard immediately and wait for the background agent to report back. Only resume execution-related work when the subagent result is available.
> **ORCHESTRATOR RULE — BACKGROUND DISPATCH**: After calling Agent() above with `run_in_background=true`, do NOT do any execution work for this phase independently. Return to the dashboard immediately and wait for the background agent to report back. Only resume execution-related work when the subagent result is available. Never call `ScheduleWakeup` or any host wake/sleep-scheduling tool while waiting (#4079) — a partial-args wake call surfaces a red validation error; the dashboard loop is the wait.
Display:

View File

@@ -409,7 +409,7 @@ Agent(
)
```
> **ORCHESTRATOR RULE — ALL RUNTIMES**: After calling Agent() above, stop working on this task immediately. Do not read more files, edit code, or run tests related to this task while the subagent is active. Wait for the subagent to return its result. This prevents duplicate work, conflicting edits, and wasted context. Only resume when the subagent result is available.
> **ORCHESTRATOR RULE — ALL RUNTIMES**: After calling Agent() above, stop working on this task immediately. Do not read more files, edit code, or run tests related to this task while the subagent is active. Wait for the subagent to return its result. This prevents duplicate work, conflicting edits, and wasted context. Only resume when the subagent result is available. Never call `ScheduleWakeup` or any host wake/sleep-scheduling tool to literalize this wait (#4079) — the Agent() call returns on its own; a partial-args wake call surfaces a red validation error.
### Handle Researcher Return
@@ -672,7 +672,7 @@ Agent(
)
```
> **ORCHESTRATOR RULE — ALL RUNTIMES**: After calling Agent() above, stop working on this task immediately. Do not read more files, edit code, or run tests related to this task while the subagent is active. Wait for the subagent to return its result. This prevents duplicate work, conflicting edits, and wasted context. Only resume when the subagent result is available.
> **ORCHESTRATOR RULE — ALL RUNTIMES**: After calling Agent() above, stop working on this task immediately. Do not read more files, edit code, or run tests related to this task while the subagent is active. Wait for the subagent to return its result. This prevents duplicate work, conflicting edits, and wasted context. Only resume when the subagent result is available. Never call `ScheduleWakeup` or any host wake/sleep-scheduling tool to literalize this wait (#4079) — the Agent() call returns on its own; a partial-args wake call surfaces a red validation error.
**Handle return:**
- **`## PATTERN MAPPING COMPLETE`:** Update `PATTERNS_PATH` to the created file path, continue to step 8.

View File

@@ -48,6 +48,15 @@ min), real, bounded call that reliably hands control back — the outer
threshold is enforced by `dispatch_ts` accumulating across calls, not by one
call's own duration.
**Never a wake-up call between cycles (#4079):** while the orchestrator waits
between `gsd_stall_watch` cycles, it must NOT call `ScheduleWakeup` (or any
host wake/sleep-scheduling tool, e.g. the `/loop` pacing surface) to
literalize "I'll wait". The `gsd_stall_watch` bash call IS the wait mechanism;
wake-up scheduling is never part of it, and a partial-args `ScheduleWakeup`
call surfaces the host's red validation error (`prompt` is required when
`stop` is not true). Just issue the next watch call (or let the blocking
Agent() return) — nothing else.
**Disclosed tradeoff:** the first cycle always sleeps a full
`PLANNER_STALL_INTERVAL_MINUTES` before its first check, so a planner that
completes in seconds is not observed by this path until that interval