chore(comments): drop the history narration main outgrew its baseline on

[CrysisDeu]

Summary

scripts/check_comment_history.py only scans the files a diff touches, so the change-history narration that landed on main in src/kiro_crew/dashboard/chat_runner.py (109 spans against a baseline entry of 103) and src/kiro_crew/context.py (12 against 11) went unnoticed on main's own pushes — and then failed any PR that touches either file with history narration in comments grew from 103 to 109 on lines that PR never wrote (first hit: #8862, round 17).

This rewords the newest spans (the issue #6189 / issue #8132 references and previously / Until #6189 phrasing in the redaction-notice docstrings of chat_runner.py; the used to be appended comment in context.py) to present-tense statements of current behavior with no issue numbers, per docs/system-specs/common/code-style.md. --write-baseline then lowers chat_runner.py to 99 (the ratchet only ever lowers). No behavior change; comments and one JSON baseline entry only.

Verification

• python3 scripts/check_comment_history.py on this branch: comment-history gate passed.
• Pristine main (fd6684b40) measured with the gate's own violations_in_source: chat_runner.py 109 vs baseline 103, context.py 12 vs 11 — the drift this PR removes.
• scripts/check_black_formatting.py and flake8 on the two files: clean.

Unblocks #8862 (and any other PR touching these two files).

[github-actions]

Design Review (Fable 5) — ✅ PASS

Design-level review of 87ea82fc64c9d554dab6d435ce56710d6ed56897 — updated in place on each push. A BLOCK verdict blocks PR readiness; PASS/CONCERNS are advisory.

Design-Verdict: PASS

Correct unblocker: rewording the drifted narration and lowering the ratchet fixes the gate failure without touching behavior or weakening any pin.

Suggestions

• The drift recurred because the gate scans only diff-touched files, so main can outgrow its baseline again and re-block unrelated PRs; a follow-up full-tree scan on main pushes would close that class permanently.

[DESIGN-REVIEWED] 87ea82f

[github-actions]

GPT 5.6 Review — ✅ no blocking findings

GPT 5.6 completed its review of 87ea82fc64c9d554dab6d435ce56710d6ed56897 and found no blocking issues.

This comment is updated in place on each push.

Review details

No findings.
[GPT-REVIEWED] 87ea82f

False positive or not applicable? A repository writer can comment:
/ai-review override gpt 87ea82fc64c9d554dab6d435ce56710d6ed56897: <one-sentence reason>

[github-actions]

First Principles Review (Fable 5) — ✅ PASS

Premise-level review of 87ea82fc64c9d554dab6d435ce56710d6ed56897 — why this exists and whether the shipped surface is the smallest honest version. Updated in place on each push. A BLOCK verdict blocks PR readiness; PASS/CONCERNS are advisory.

All verification is done. The gate script, code-style rule, and audit workflow confirm the description's mechanism; I counted the removed spans in the patch (10 in chat_runner.py: 109→99 matching the new baseline entry; 1 in context.py: 12→11 matching its untouched entry of 11), and the baseline lowering is itself mandated by the gate's shrink rule. No pins are deleted; the "why" content of each docstring survives in present tense. Final review:

First-Principles-Verdict: PASS

Verify whole-tree scope once: run RATCHET_SCOPE_WHOLE_TREE=1 python3 scripts/check_comment_history.py on the merge — the PR's own diff-scoped run cannot see any third drifted file.

What this change ships

Inventory (3 items)

Intent: unblock PRs failing the comment-history gate on drift they never wrote, by removing main's over-baseline narration. FIX.

1. Redaction-notice docstrings in chat_runner.py now state current behavior, no issue numbers or past tense — justified
2. The context.py prefix-channel comment states current behavior instead of "used to be appended" — justified
3. The chat_runner.py baseline entry drops from 103 to 99 — justified

The defect is derived, not asserted: the gate's grown-count verdict (scripts/check_comment_history.py:401) fires on any PR touching a file whose main count exceeds baseline, and I re-counted the patch's removed spans (10 and 1) against the gate's own patterns — the arithmetic in the description checks out. The rewording is mandated by docs/system-specs/common/code-style.md ("Comments explain the WHY"); the baseline lowering is mandated by the gate's must-lower rule, so neither is a rider. Recurrence is already owned by main-ratchet-audit.yml, which runs this gate whole-tree on every main push — no new mechanism needed, and none added.

[FIRST-PRINCIPLES-REVIEWED] 87ea82f

[github-actions]

Opus 4.8 Review — ✅ no blocking findings

Reviewed 87ea82fc64c9d554dab6d435ce56710d6ed56897 — this comment is updated in place on each push.

Review details

No findings.

[OPUS-REVIEWED] 87ea82f

Verdict parsed from the review's SHA-scoped output markers for commit 87ea82fc64c9d554dab6d435ce56710d6ed56897.

False positive or not applicable? A repository writer can comment:
/ai-review override fable 87ea82fc64c9d554dab6d435ce56710d6ed56897: <one-sentence reason>