docs(autonomy): revise the semi definition across the protocol and docs (#2253 Bolt semi-docs-revision) - #2321
Merged
Merged
Conversation
|
Important Review skippedReview was skipped as selected files did not have any reviewable changes. 💤 Files selected but had no reviewable changes (21)
⛔ Files ignored due to path filters (197)
⚙️ Run configurationConfiguration used: Path: .coderabbit.yaml Review profile: ASSERTIVE Plan: Pro Plus Run ID: ⛔ Files ignored due to path filters (197)
📒 Files selected for processing (21)
You can disable this status message by setting the Use the checkbox below for a quick retry:
Comment |
![cursor[bot]](https://avatars.githubusercontent.com/in/1210556?s=60&v=4){kind=small}
There was a problem hiding this comment.
Cursor Bugbot has reviewed your changes using default effort and found 1 potential issue.
❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, enable autofix in the Cursor dashboard.
Reviewed by Cursor Bugbot for commit ad24560. Configure here.
| "receivedAt": "2026-08-05T08:30:06Z", | ||
| "detail": "ballot accepted: subagent-1", | ||
| "voter": "subagent-1" | ||
| }, |
There was a problem hiding this comment.
Ballot submitted after acceptance
Low Severity
Election timeline records ballot acceptance for subagent-1 at 2026-08-05T08:30:06Z while the ballot’s submittedAt is 2026-08-05T09:55:00Z, so submission time precedes receipt in the ledger.
Additional Locations (1)
Reviewed by Cursor Bugbot for commit ad24560. Configure here.
j5ik2o
force-pushed
the
bolt-semi-docs-revision
branch
from
August 6, 2026 00:49
ad24560 to
3a67f3c
Compare
…ap (user ruling)
…nce, occurrence mapping)
…nator, signature drift, verbatim cites)
…roduction wiring, edge strength)
…with reviewer sign-off
… with reviewer sign-off
…nit with reviewer sign-off
…unit with reviewer sign-off
…with reviewer sign-off
…nings - semi-docs-revision FD authored (docs 64-line R/P/U inventory, stage-protocol 9-line dispositions), reviewer READY iteration 1 - stop-question-carveout FD repaired: missing S12a review detected on session handover, iteration 1 BLOCKERs (unmeasured line-range citations) fixed with measured verbatims, iteration 2 READY - S13 election E-SRA-FDS13 (2-0): adopted L1 (static contract ACs must bind verification means) and L2 (c3-mps FD-side 3-point wiring specify) into project.md
- security-design + logical-components authored per unit (kind-resolved produces; spec unit ships security-design only) - S12a reviews: 5 units READY iteration 1, autonomy-statusline READY iteration 2 (NFR-3 classification fixed), semi-docs-revision NOT-READY at limit with the single residual (one bare BR reference formatting) accepted under E-LSSADS13 mechanical-class rule, closure grep recorded in diary - S13 election E-SRA-NDS13 (2-0, GoA 1x2): zero adoptions
FR-DISP-1: the statusline names the active Intent's autonomy mode with the same vocabulary `--status` prints (none / semi / full), so users read one word for the autonomy level instead of inferring it. The decision is a pure function `autonomySegment` in amadeus-lib.ts: it reads `Intent Autonomy Mode` out of the state text the hook has already loaded (no extra I/O) and returns the bare mode name, or "" when the field is absent or outside the canonical AutonomyMode domain. The hook keeps a single wiring line in the active-workflow branch; the three early returns are untouched. The mode domain is pinned to the canonical AutonomyMode union through a type-only import, so widening it there fails typecheck here instead of diverging silently. t448 imports the function from the shipped surface and fixes the whole decision table: three canonical modes render, absent field and out-of-domain values degrade to no segment.
The segment's conditional raised main's cyclomatic complexity from 26 to 27, and the committed complexity baseline only ratchets down. Extracting the concatenation into a named helper returns main to CCN 26 while the segment's own branch and the bare-mode-name return domain are preserved verbatim inside the helper. Named rather than anonymous so the file's anonymous-function ordinals stay stable — a shifted ordinal would report an untouched function as a new violation. Behaviour is unchanged: the segment still renders only on the active-workflow line, and the three early returns are untouched. Declared refinement of the unit's functional design, which scoped the wiring to a single line in main; the ruling authorises the extraction.
…cord batch 1 rulings - tNNN reservation collided with the advanced base (max t447): mechanical renumber across 33 record files (core->t451-453, policy->t454/455, stop->t456, flag->t449/450, advisory->t457-459; statusline keeps t448), zero residuals - code-generation diary: isolation blocker + user ruling (isolated-worktree builders with ff adoption), E-SRA-CG1 deviation election (B: named helper extraction), mechanical application of the same ruling to the parseNextFlags pre-pass (A) - elections: E-SRA-CG1 store
…kpoint audit shard
…rd the recovery merge resolution
Revise the semi definition to the new semantics: questions are resolved unattended through the same five-rung ladder as full (recorded as AUTO_DECIDED, last two rungs entering the unreviewed queue), while milestones (phase boundaries, walking skeleton, Intent completion) still require a human. semi holds no Intent grant, accepts --policies-file as confirmed-policy material, and takes effect only for human-command provenance. Sync the mode recording bullet with the --autonomy launch flag contract. Walking-skeleton human-wait lines (:105/:808) preserved. Refs #2253
…human milestones Rewrite every docs line that stated the old semi definition (questions human-owned) to the new semantics: questions auto-decided through the same resolution ladder as full, humans retained only at milestones (phase boundaries, walking skeleton, Intent completion). Split the stop hook suppression description: the question-tag carve-out is suppressed under full and human-command semi, the conversational carve-out under full only. English and Japanese pairs updated together; walking-skeleton human-wait lines preserved unchanged. Refs #2253
j5ik2o
force-pushed
the
bolt-semi-docs-revision
branch
from
August 6, 2026 02:31
f0cd275 to
460fe99
Compare
j5ik2o
marked this pull request as ready for review
August 6, 2026 02:45
j5ik2o
added a commit
that referenced
this pull request
Aug 6, 2026
… (#2372) Bring the workflow record for #2253 onto main: the build-and-test stage's seven artifacts, the construction phase-boundary check, the goal reconciliation items and receipt, the appended audit shard, and the terminal state. The registry row advances in-flight → complete. Implementation for this intent already landed in seven PRs (#2293, #2294, #2295, #2316, #2317, #2318, #2321); this commit carries the record only — no source or test changes. Evidence recorded in the artifacts (measured on the conductor clone, values transcribed from command output): - bash tests/run-tests.sh --ci: exit 0, RESULT: PASS (11,494 pass / 0 fail) - build / typecheck / lint / source-only:check / complexity-gate: exit 0 - declared sensors on the seven artifacts: FIRED 14 / PASSED 14 / FAILED 0 - requirement traceability: 33 of 33 ids trace into the construction tree - formal-model-check: NOT_DETECTED (correlated run, no counterexample) - goal reconciliation: goal-statement ACHIEVED over 14 digest-verified refs The audit shard is a pure append over main's copy (main's bytes verified as a byte-exact prefix); the registry keeps every other row untouched. Refs #2253
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
概要
Intent
260805-semi-redefine-autonomy-f(Issue #2253)の Construction Boltsemi-docs-revision(swarm batch 4、docs 専任)。semi 再定義(「full − 節目の自動裁定」)の文書面を canonical と docs へ反映する。packages/framework/core/amadeus-common/protocols/stage-protocol.mdの 4 行のみ改訂(:33 人間裁定列挙の直接反転 / :119 / :125--autonomy none|semi|full追記 / :131 新 semi 定義 8 要素)— 保存対象 5 行は diff 非出現を機械確認マージ順制約(stacked PR)
本 PR は conductor ブランチ(batch 1+2 の全実装を含む)を base に持つ。挙動 Unit の PR(#2293 / #2294 / #2295 / #2316 / #2317 / #2318)がすべて着地した後、最後にマージすること — docs は挙動着地後に出荷する(FD domain-entities の Unit 間関係)。それまで diff に先行 Bolt のコミットが含まれる。
検証(builder 実測)
V1 旧定義残存 grep 0 / V2 禁止語彙 0 / V3 保存行 diff 非出現 / V4 対訳 8 ペア閉包 / V5 build 後 tracked 不変 / V6 = 24(増分理由は棚卸しで確定)/ typecheck 0 / lint 0 / full
bash tests/run-tests.sh --ciRESULT: PASS。referee converged / finalize(batch 4)converged。既知の外部ブロッカー
「No silent drop (trusted base ratchet)」は main 側 reconcile の連続失敗(#2313)により本 PR と無関係に赤になる。#2313 解消待ち。
Refs #2253
Note
Low Risk
Changes are documentation, reverse-engineering notes, and election metadata under
amadeus/spaces/default/with nopackages/or test code in the diff; risk is misalignment between PR description and actual files merged, not production behavior.Overview
This PR does not touch runtime code in the supplied diff. It updates workflow record artifacts for Issue #2253 (
semiredefinition and launch autonomy).Code knowledge base (
amadeus/spaces/default/codekb/amadeus/) — Promotes260805-semi-redefine-autonomy-fto the “current” narrative across shared RE docs (api-documentation,architecture,business-overview,code-quality-assessment,code-structure,component-inventory,dependencies,technology-stack,reverse-engineering-timestamp). New material documents observed behavior at2f255bc69: howsemicurrently only auto-approves phase-internal stage gates, the five-stepresolveAutoDecisionladder (vs. “4 steps”),--autonomyabsence in code,set-autonomy --policies-filesilent drop for non-full, stop-hook cap vs. question carve-out asymmetry, bolt 8 autonomy subcommands, and pinned tests (t431,t121) that a redefinition would flip. The prior260805-pr-convergence-plugin“current” sections are removed or demoted to history (notably inarchitecture,code-quality-assessment,code-structure,component-inventory).Re-scan — Adds
re-scans/260805-semi-redefine-autonomy-f.mdas the authoritative differential scan (findings, line corrections vs. developer scan, sync surfaces: 9semilines in canonicalstage-protocol.md, 22 doc files, test inventory).Elections — Introduces full election trees under
amadeus/spaces/default/elections/(ballots, ledger, tally, timeline, views,record.md), including among others: E-SRA-RA1 (batch adoption of requirements-analysis Q1–Q6 recommendations), E-SRA-ADS13 (persist conductor re-verification when reviewers cannot read out-of-scope code), E-SRA-CG1 (helper extraction vs. complexity baseline for statusline C14 wiring), E-SRA-DPS13 / FDS13 / NDS13 (§13 learning adoption votes).Note: The PR title/description references canonical
stage-protocol.mdand docs translation pairs; those paths are not present in this diff—only record/codekb and election JSON/Markdown appear here (consistent with a stacked branch or a separate bolt).Reviewed by Cursor Bugbot for commit ad24560. Configure here.