docs: record the ninth 2026-09-04 block and number the SC-10 parked PRs

[Chris0Jeky]

Summary

Two docs-only edits.

1. Adds the ninth 2026-09-04 block to docs/STATUS.md, recording the remainder of the
   overnight orchestrator lane: main c174a517a to 9943804bc, nine merges. Same shape as the
   seventh and eighth blocks — one bullet per merged PR with issue, PR, merge SHA, shipped
   behaviour, evidence counts and what is not verified, grouped capture composer / review and
   board frontend / backend / harness-CI-docs, closing with a "Not shipped reality" line.
   The header states once that Codex credits stayed exhausted (SC-9) so every PR carried one
   fresh-context independent reviewer; that control-plane PRs #2502, #2506, #2522 and
   #2532 are parked for the maintainer under ADR-0066 / SC-10; and the stacked-base trap —
   a PR stacked on a merged-but-undeleted base branch merges into that branch, not main, so
   re-target with gh pr edit N --base main first.
2. Amends the SC-10 human-action item in OUTSTANDING_TASKS.md §J so the parked control-plane
   PRs are all named by number.

Recorded PRs and merge SHAs

PR	Issue	Merge
#2527	#2141 (Refs)	ca6fe42fe
#2530	#2434 (re-target of #2496)	b83dc6e11
#2529	docs (eighth block)	b461be49f
#2533	#2400	5aa792707
#2528	#2465	94d2b8217
#2500	#1700	dffb01de8
#2534	#2198	d11bd4ada
#2536	#2472	298eb5444
#2539	#2490	9943804bc

Two things recorded rather than silently omitted:

• The eighth block's "#2496 did not reach main" note is now resolved. #2530 carries
  #2496's exact delta re-targeted onto main; c60d0156a is still not an ancestor of main.
  Follow-up #2523.
• #2536's reviewer caveat is carried into the block. The stated mechanism (a stale
  /root/.npm cache persisting across builds) is not supported by Docker layer semantics, so the
  root cause of the cacache race stays unconfirmed; the change is still strictly narrower and its
  ledger row is honestly open.

Other follow-ups named in the block: #2537 (#2533), #2538 (#2500), #2540 (#2539).
#2500 was itself re-targeted off a merged stacked base before merging — the trap, avoided.

The SC-10 edit

One line in OUTSTANDING_TASKS.md §J: "the #1898 dev-up port-release PR once open" becomes
"#2522 (dev-up port release, #1898)", and "#2532 (release-cache scanner comment handling,
CI-09)" is added to the parked list. This closes the non-blocking MEDIUM declined on #2529's
review gate. Nothing is ticked; no other text changed.

Docs-only; SHAs read from GitHub (gh pr view --json mergeCommit) and cross-checked against
git log --first-parent c174a517a..origin/main. No masterplan or ADR edits; Last Updated
unchanged at 2026-09-04.

Verification

• node scripts/check-docs-governance.mjs — passed.
• git diff --stat — docs/STATUS.md +20, OUTSTANDING_TASKS.md +1/-1; no reflow of
  neighbouring blocks. git diff --check clean.
• The b461be49f CI claim was measured, not transcribed:
  gh api repos/.../commits/b461be49f.../check-runs returns 17 checks, all completed and
  success except Secret Scan (skipped).
• Follow-up issues #2523, #2537, #2538, #2540 confirmed OPEN with gh issue view.

Not verified: the per-PR evidence counts are transcribed from each PR body and its review-gate
comment; no backend, frontend or Playwright suite was re-run for this docs change.

[chatgpt-codex-connector]

You have reached your Codex usage limits for code reviews. You can see your limits in the Codex usage dashboard.
To continue using code reviews, add credits to your account and enable them for code reviews in your settings.

[Chris0Jeky]

Review gate (docs-only, no connector): one fresh-context reviewer verified all nine behavioural claims TRUE against the tree at the head (including the #2434 fix now on main, the strip-then-truncate ordering in all three log wrappers, and the permission-mode change framed as a test/model change), the SC-10 line names #2502 #2506 #2522 #2532 + the #2331 proposal and records no decision, and the 'Not shipped reality' line does not overclaim. The two GitHub-state assertions (17 checks at b461be4; #2522/#2532 fresh-context reviewed) are true per the review-gate comments on those PRs. Verdict SHIP. Merging when the docs gate is green.