Repository navigation
docs(rfc): withdraw the host_pressure record, wake and pause; keep the critical refusal - #37170
Conversation
UX Review (Opus 5.5) — ⏭️ skippedRevision |
Design Review (Opus 5.5) — ✅ PASSDesign-level review of Design-Verdict: PASS This PR only changes documentation: it records the owner's withdrawal, removes nothing that code reads, and leaves the withdrawn text in place as the record. Notes
[DESIGN-REVIEWED] c73b40a Description read[DESCRIPTION-READ] d74b35931cf88aa066d5b21eda910a42e7af76677311d37067c2c242968bb277 That sha256 names everything this verdict read: the PR description, plus the 1 evidence file(s) the workflow collected from it and handed this review. It covers that evidence because the media strip replaces every attachment URL with the same placeholder, so an edit swapping one attachment for another would leave the description's captured text identical while what this review judged changed. Recompute it with |
First Principles Review (Opus 5.5) — ⏭️ skippedRevision |
…t pressure The repository owner narrowed rfc-work-ledger-resource-pressure on 2026-11-14: the conductor checks resource_status before it dispatches, and only one piece of code stays. This removes the rest of #36481: - src/kiro_crew/host_pressure.py (the host record) and its RESERVED_ROOT_DIRS slot - ResourcePressureNotifier's episode id, pressure_sink, level-change wake and pending flush (back to its pre-#36481 behaviour) - the pressure wake in conductor_wake and WorkLedgerProbe - conductor.host_pressure in work_ledger_read - the host_pressure conductor rule in spec_prompts, crewmate_prompts.json and goal-conductor SKILL.md, its parity anchor, and its work-ledger.md section - test/test_work_ledger_host_pressure.py; the golden and spec-byte pins go back to their pre-#36481 values Kept: _refuse_if_host_critical in _create_item (409 via the live admission_check). Its code is renamed host_pressure_critical -> host_memory_critical, since nothing named host_pressure is left. Its tests move to test/test_work_ledger_host_memory_refusal.py. The withdrawal itself is recorded in the RFC by #37170, which lands first.
Opus 5.5 Review — ✅ no blocking findingsReviewed Review detailsNo findings. Notes:
[OPUS-REVIEWED] c73b40a Verdict parsed from the review's SHA-scoped output markers for commit False positive or not applicable? A repository writer can comment: |
GPT 6.1 Review — ✅ no blocking findingsGPT 6.1 completed its review of This comment is updated in place on each push. Review detailsNo findings. False positive or not applicable? A repository writer can comment: |
Code change approver recommendation: ApproveWhy: the decider estimates a 7% chance that this change introduces a defect, within the 12% limit for approving on its own, and every check passes. Docs-only RFC status change recording the owner's withdrawal of the host_pressure record, wake, pause rule and §12, keeping only the critical create refusal. Claims check out against main: _refuse_if_host_critical exists and is called from _create_item. An automated recommendation, not a review: a maintainer still decides. It approves on its own only when the estimated chance of a defect is at most 12% and the change has at most 400 changed lines. Decider v3c (217fe607) · commit |
b03d169 to
7fe600e
Compare
…t pressure The repository owner narrowed rfc-work-ledger-resource-pressure on 2026-11-14: the conductor checks resource_status before it dispatches, and only one piece of code stays. This removes the rest of #36481: - src/kiro_crew/host_pressure.py (the host record) and its RESERVED_ROOT_DIRS slot - ResourcePressureNotifier's episode id, pressure_sink, level-change wake and pending flush (back to its pre-#36481 behaviour) - the pressure wake in conductor_wake and WorkLedgerProbe - conductor.host_pressure in work_ledger_read - the host_pressure conductor rule in spec_prompts, crewmate_prompts.json and goal-conductor SKILL.md, its parity anchor, and its work-ledger.md section - test/test_work_ledger_host_pressure.py; the golden and spec-byte pins go back to their pre-#36481 values Kept: _refuse_if_host_critical in _create_item (409 via the live admission_check). Its code is renamed host_pressure_critical -> host_memory_critical, since nothing named host_pressure is left. Its tests move to test/test_work_ledger_host_memory_refusal.py. The withdrawal itself is recorded in the RFC by #37170, which lands first.
7fe600e to
701b8c1
Compare
701b8c1 to
0efcbcf
Compare
…e critical refusal
0efcbcf to
c73b40a
Compare
|
This PR's Goal records a decision by the repository owner: withdraw the host_pressure record, the wake and the pause rule, and keep only the critical Review result on this head: no blocking findings. The change touches docs only, in 3 files under docs/request-for-change/. PR Readiness is green. Each claim checks out on main. For the security review: nothing to note. The change keeps the critical refusal and widens no security, sandbox or approval gate. Optional:
|
|
Simplify the design on resource detection while dispatching worker |
…t pressure The repository owner narrowed rfc-work-ledger-resource-pressure on 2026-11-14: the conductor checks resource_status before it dispatches, and only one piece of code stays. This removes the rest of #36481: - src/kiro_crew/host_pressure.py (the host record) and its RESERVED_ROOT_DIRS slot - ResourcePressureNotifier's episode id, pressure_sink, level-change wake and pending flush (back to its pre-#36481 behaviour) - the pressure wake in conductor_wake and WorkLedgerProbe - conductor.host_pressure in work_ledger_read - the host_pressure conductor rule in spec_prompts, crewmate_prompts.json and goal-conductor SKILL.md, its parity anchor, and its work-ledger.md section - test/test_work_ledger_host_pressure.py; the golden and spec-byte pins go back to their pre-#36481 values Kept: _refuse_if_host_critical in _create_item (409 via the live admission_check). Its code is renamed host_pressure_critical -> host_memory_critical, since nothing named host_pressure is left. Its tests move to test/test_work_ledger_host_memory_refusal.py. The withdrawal itself is recorded in the RFC by #37170, which lands first.
…t pressure (#37164) The repository owner narrowed rfc-work-ledger-resource-pressure on 2026-11-14: the conductor checks resource_status before it dispatches, and only one piece of code stays. This removes the rest of #36481: - src/kiro_crew/host_pressure.py (the host record) and its RESERVED_ROOT_DIRS slot - ResourcePressureNotifier's episode id, pressure_sink, level-change wake and pending flush (back to its pre-#36481 behaviour) - the pressure wake in conductor_wake and WorkLedgerProbe - conductor.host_pressure in work_ledger_read - the host_pressure conductor rule in spec_prompts, crewmate_prompts.json and goal-conductor SKILL.md, its parity anchor, and its work-ledger.md section - test/test_work_ledger_host_pressure.py; the golden and spec-byte pins go back to their pre-#36481 values Kept: _refuse_if_host_critical in _create_item (409 via the live admission_check). Its code is renamed host_pressure_critical -> host_memory_critical, since nothing named host_pressure is left. Its tests move to test/test_work_ledger_host_memory_refusal.py. The withdrawal itself is recorded in the RFC by #37170, which lands first. Co-authored-by: Joe Guo <zejiangg@amazon.com>
Problem / Motivation
Goal: Record on main that the repository owner withdrew the host-pressure record, conductor wake and pause rule, and kept only the critical
createrefusal.The repository owner, verbatim: "我的想法是缩小,其实只要给个简单指示,说明派活要先查资源,然后给AI工具查资源不就好了?为什么改那么多代码" and "开精简 PR,只留 critical 拒绝。" In English: the conductor only needs a short rule to check resources before it dispatches, and the
resource_statustool already does that; keep only the critical refusal.Why it matters
The code trim #37164 removes a conductor behaviour that rfc-work-ledger-resource-pressure.md on main still records as decided. The First Principles lane reads that decision from the base branch, so the withdrawal must be on main first, in its own PR, before the trim can pass.
Not a goal
What changed (motivation → approach → change)
rfc-work-ledger-resource-pressure.md: the status header records the owner's withdrawal of §3.1 to §3.4, the tight advice in §3.5 and the §12 amendment, with the owner's words; only the §3.5 critical refusal stays. Statusin-progress, with refactor(work-ledger): keep only the critical create refusal from host pressure #37164 as the implementation that removes the rest. §12 is marked withdrawn (#36657 closed unmerged).rfc-work-ledger-person-wait-hold.md: the "Amended by §12" line from docs(rfc): a host_pressure level change releases a person-wait hold once #36589 now says that amendment was withdrawn and never built.README.md: the status row says the same.Backwards compatibility
Compatible: docs only; no code, config or behaviour changes here.
Removes nothing: the withdrawn design text stays in the RFC as the record; only its status lines change.
Tests
Proof: not provable -- docs or config only
docs-lint passes locally.
Manual verification
N/A — docs only.
Related Issues
Part of #36301
Checklist
feat|fix|docs|style|refactor|perf|test|chore|ci|build|revert: ...)