Session Handoff [review-tooling]: Tag-Opener Primitive Released as 2.0.1089, TAGS Stripper Next #3211

Next steps, in priority order

1. Strip an Attributed Details or Summary Opener Through TAGS in pr_review.py #3208: strip an attributed or self-closed <details> or <summary> opener through TAGS in pr_review.py. TAGS is the one opener reader left off tag_opener, so <details open> and <summary class="x"> print raw from unwrap and stay on an entry_heading read. Done looks like: TAGS ends a tag name where tag_opener does, bounded so an unclosed tag strips nothing past its own line, with a test pinning <details open>, <summary class="x">, and <summary/> as stripped and <summaryx> as kept.
2. pr_review.py Passes a Clean-Verdict Copilot Round Whose Headline Sentence Names a Finding #2567: a clean-verdict round whose headline names a finding. Needs the maintainer's design call on reading headline prose before any code, so ask before starting it. It carries no decision label, so raise it as a question when it comes up.
3. Finish the Incidental Count Sweep, Including Counted Section Titles #3169: finish the incidental-count sweep. Carried over, lower priority.

External blockers

• CodeRabbit's free OSS review allowance ran out again on promotion Promote the Details and Summary Opener Primitive in pr_review.py to Main #3210, after reviewing feature Recognize a Details or Summary Opener Through One Primitive in pr_review.py #3209 clean. It is advisory and blocks nothing, so proceed on Copilot and the local pass when its notice says the limit is reached, and do not ask it again on that pull request.

Internal dependencies

• Strip an Attributed Details or Summary Opener Through TAGS in pr_review.py #3208 builds on tag_opener, which is released, so it starts from a current develop and needs no ordering against anything else.
• pr_review.py Passes a Clean-Verdict Copilot Round Whose Headline Sentence Names a Finding #2567 waits on the maintainer's design call.

State

• Released 2.0.1089 from promotion Promote the Details and Summary Opener Primitive in pr_review.py to Main #3210 (Recognize a Details or Summary Opener Through One Primitive in pr_review.py #3209, closing Recognize a Details or Summary Opener Through One Primitive in pr_review.py #3206).
• This machine's Codex and opencode skills snapshot was refreshed from main at d2e732d8 and reads current. The host-safety kit reports CURRENT.
• The primary checkout is on develop, clean, and at origin/develop. pr_review.py cleanup removed the task worktree and branch, locally and on the remote.
• Detached fm- worktrees belong to another session. cleanup --sweep reports the fm-401k-promotion and fm-main ones as removable, being idle past their window and contained in main. --apply was not run, since it reaches that session's trees, so it is the maintainer's call.
• No other repository was touched.

The parked decision queue

The queue held 0 open issues with the decision label when this was written. Nothing is parked.

What the last round did

• Recognize a Details or Summary Opener Through One Primitive in pr_review.py #3209 adds tag_opener, one tag-name boundary on CommonMark's rule (HTML whitespace, the end of a line, >, or />), and moves DETAILS_OPEN, SUMMARY, DETAILS_OPENER, HEADING, DETAILS_TAG, FOLD_OPENER, FOLD_STOP, and heading_of onto it. heading_of now reads its heading through one anchored SUMMARY.match, so a block whose own summary is rejected no longer takes a nested entry's summary. Closes Recognize a Details or Summary Opener Through One Primitive in pr_review.py #3206 via Promote the Details and Summary Opener Primitive in pr_review.py to Main #3210.
• Filed Strip an Attributed Details or Summary Opener Through TAGS in pr_review.py #3208 from the local review of Recognize a Details or Summary Opener Through One Primitive in pr_review.py #3209.

What not to repeat

• HTML's tag-name rule is not the one that matters here. The first cut ended a name on a bare /, as the HTML tokenizer does, and an existing test failed: GitHub renders Markdown, and CommonMark hands HTML no tag for <details/x>. The boundary is CommonMark's.
• A mutation check must revert to the pre-branch code, not to an earlier cut of the branch. One check reverted heading_of's first guard to the branch's own intermediate SUMMARY.match rather than the original startswith, so the test it "proved" passed against the real base. The fourth local pass caught it.
• A docstring claiming "every reader" draws a finding per exception. Naming TAGS as the exception drew RATE_LIMITED_REPLY next. Scoping the claim by property ("every opener reader that admits an attribute") ended it.
• The local pass took five rounds on a small diff. Rounds two through four each found defects in the previous round's fix text, the pattern local-strict-review's two-round budget exists for.

New learnings

• None beyond the issue filed and the points above.