Session Handoff [engineering]: #2430 and #2750 Released as 2.0.972, #2839 and #2838 Next #2841

Next steps, in priority order

1. pr_review.py wait Matches a Required Check by Name Alone, Ignoring the Rule's integration_id #2839, pr_review.py wait matches a required check by name alone. The ruleset entry carries an integration_id, and a same-name check from another app satisfies the hold today. Done means a rollup node satisfies an entry with an integration_id only when its app matches, with a test for the same-name, different-app case.
2. Sentence-length comment and docstring readers miss four layouts #2838, sentence-length gaps left past the edit budget on Extend the Sentence-Length Check to Comments and Docstrings #2836. Indented tables under a colon line in comments, NumPy docstring sections, Google Args: bodies, bullet and numbered markers counted as words, a closing quote on a docstring's last code line, and docstring boundaries inferred from quotes. Done means each listed case is handled or explicitly ruled out on the issue.

#1814 (skill files hard-wrap) is still open and is left to the unattended loop's picker. #2830 tracks a diff-floor test that fails on a clean develop when run from some working directories, and CI passes it.

External blockers

• Nothing in the next steps waits on the maintainer.
• CodeRabbit still posts only its low-star notice here. Copilot plus the attested local pass covered this round.

Internal dependencies

• None among the next steps. pr_review.py wait Matches a Required Check by Name Alone, Ignoring the Rule's integration_id #2839 and Sentence-length comment and docstring readers miss four layouts #2838 touch disjoint files (scripts/pr_review.py and .github/actions/prose-gate/prose_lint.py), so they can run as parallel lanes again.

State

• Extend the Sentence-Length Check to Comments and Docstrings #2836 and Hold an Attested Wait Until the Base Branch's Required Checks Register #2837 merged to develop, Promote the Comment and Docstring Sentence-Length Check and the Attested Wait Hold to Main #2840 promoted them to main, and it was released as 2.0.972. Re-derive both.
• The primary checkout is on develop, fast-forwarded. Both lane worktrees, their local and remote branches, and the detached skills-refresh worktree are removed.
• The installed Skills snapshot reads current at the promoted main, and the host-safety kit reads CURRENT.
• The decision queue held zero open issues when this link was written.
• Other tracks' open handoffs are untouched: Session Handoff [windows]: #2578 Landed for #694, Closing Through Promotion #2580, #1171 Next #2581 (windows), Session Handoff [host-instructions]: Hardening Rule and Merge Guard Released as 2.0.867 and 2.0.868, #2499 Docstring Next #2514 (host-instructions), Session Handoff [default]: Four Follow-Ups Landed and Released as 2.0.826, Downstream Adoption Next #2361 (default), Session Handoff [python-contract]: Escaped and Folded python-directories Values Refused, Released as 2.0.738 #2014 (python-contract).

The parked decision queue

Zero open decision issues.

What the last round did

• prose_lint's sentence-length Check Skips Code Comments Although the 25-Word Cap Covers Them #2430 landed as Extend the Sentence-Length Check to Comments and Docstrings #2836. prose_lint.py's opt-in sentence-length check now reads code comments outside Markdown and Python docstrings, through one counting helper shared with the Markdown reader, scoped to changed lines. The comment prose filters are now shared with comment-wrap.
• pr_review.py wait Exits 0 on an Attested Fix Push Before Its Checks Register #2750 landed as Hold an Attested Wait Until the Base Branch's Required Checks Register #2837. On a BLOCKED merge, wait holds an attested head while a check the base branch's rules require is absent from the rollup. An unreadable rules read keeps the old behavior and prints a note. attest waits up to 60 seconds for a push GitHub has not registered, only where the remote branch already holds the checkout's commit.
• Both ran as parallel standard lanes from one shared brief plus a per-lane design ruling, then one promotion, Promote the Comment and Docstring Sentence-Length Check and the Attested Wait Hold to Main #2840. Its only finding was the pr_review.py wait Matches a Required Check by Name Alone, Ignoring the Rule's integration_id #2839 gap, answered as a deferral.
• Filed Sentence-length comment and docstring readers miss four layouts #2838 and pr_review.py wait Matches a Required Check by Name Alone, Ignoring the Rule's integration_id #2839. A decline reply on Hold an Attested Wait Until the Base Branch's Required Checks Register #2837 claimed a same-name check would still end in CHECKS_PENDING, which the code does not guarantee. A correction was posted under it, pointing at pr_review.py wait Matches a Required Check by Name Alone, Ignoring the Rule's integration_id #2839.

What not to repeat

• A worker's decline reply that overstates what the code guarantees. Read each decline a lane reports for claims beyond the design ruling, and correct it on the thread before the promotion opens, since the promotion's reviewer raises the same gap again.
• Lane 2430 saw pr_review.py wait return at once while only a Windows unit-test check was pending. It ran the pre-Hold an Attested Wait Until the Base Branch's Required Checks Register #2837 script from its own checkout. Check whether Hold an Attested Wait Until the Base Branch's Required Checks Register #2837's required-check hold covers that case before filing it, since the required aggregator waits behind that job.

New learnings

• Nothing new that is durable. The entries above are specific to this round's attempts.