Session Handoff [review-tooling]: Attributed Opener Stripper Released as 2.0.1090, #2567 Design Call Next #3217

Next steps, in priority order

1. 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. This waits on the maintainer's design call, which was put to the maintainer at the close of this round. The answer is in this issue's comments, or on pr_review.py Passes a Clean-Verdict Copilot Round Whose Headline Sentence Names a Finding #2567 itself. Done looks like the arm pr_review.py Passes a Clean-Verdict Copilot Round Whose Headline Sentence Names a Finding #2567 states: closed as not worth a prose reader, or a reader with both tests pr_review.py Passes a Clean-Verdict Copilot Round Whose Headline Sentence Names a Finding #2567 names.
2. End a Code Span at an HTML Block Line in pr_review.py #3214: end a CODE_SPAN at an HTML block line. unwrap now masks through CODE_SPAN, so a stray backtick that pairs across </details>/<details> leaves those wrappers printed raw. Done looks like the constructed case in End a Code Span at an HTML Block Line in pr_review.py #3214 stripped by unwrap, with every other CODE_SPAN reader's tests still green.
3. Share One Attribute Tail Across the Details and Summary Opener Readers in pr_review.py #3212: one attribute tail for the remaining opener readers. TAGS and HTML_TAG now share ATTRIBUTES. SUMMARY, DETAILS_OPENER, and DETAILS_TAG each still spell their own tail. Done looks like a table test pinning the same opener shapes at every site.
4. Keep an Emphasis Tag Quoted in Code When reduced Strips Emphasis in pr_review.py #3215: keep an emphasis tag quoted in code when reduced strips emphasis. It is shared by normal, unwrap, and heading_count, so the vetted, printed, and counted text must move together.
5. Finish the Incidental Count Sweep, Including Counted Section Titles #3169: finish the incidental-count sweep. Carried over, lowest priority.

External blockers

• pr_review.py Passes a Clean-Verdict Copilot Round Whose Headline Sentence Names a Finding #2567 waits on the maintainer's design call, per step 1. Steps 2 through 5 do not wait on it.
• CodeRabbit does not review this repository automatically, since it has fewer than 10 stars. Copilot and the local pass carried Strip an Attributed Details or Summary Opener Through TAGS in pr_review.py #3213 and Promote the Attributed Opener Stripper in pr_review.py to Main #3216.

Internal dependencies

• End a Code Span at an HTML Block Line in pr_review.py #3214 before Keep an Emphasis Tag Quoted in Code When reduced Strips Emphasis in pr_review.py #3215. Both change what unwrap masks and prints, and End a Code Span at an HTML Block Line in pr_review.py #3214 changes which spans are masked, so Keep an Emphasis Tag Quoted in Code When reduced Strips Emphasis in pr_review.py #3215's fix is easier to test on the corrected spans.
• Share One Attribute Tail Across the Details and Summary Opener Readers in pr_review.py #3212 is independent of the others, but touches SUMMARY, which heading_of reads, so run the whole test_pr_review suite rather than the opener table alone.

State

• Released 2.0.1090 from promotion Promote the Attributed Opener Stripper in pr_review.py to Main #3216 (Strip an Attributed Details or Summary Opener Through TAGS in pr_review.py #3213, closing Strip an Attributed Details or Summary Opener Through TAGS in pr_review.py #3208).
• This machine's Codex and opencode skills snapshot was refreshed from main at 07b0a520 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, and --apply was not run, since it reaches that session's trees.
• No other repository was touched.

The parked decision queue

The queue held 0 open issues with the decision label when this was written. #2567 carries no decision label and was put to the maintainer directly at the close of this round, per step 1.

What the last round did

• Strip an Attributed Details or Summary Opener Through TAGS in pr_review.py #3213 makes TAGS strip an attributed or self-closed <details> or <summary> opener. The attribute tail is HTML_TAG's grammar, now shared as ATTRIBUTES. unwrap finds its tags in a copy masked by mask_quotations, so a tag a finding quotes in code stays in its text. Closes Strip an Attributed Details or Summary Opener Through TAGS in pr_review.py #3208 via Promote the Attributed Opener Stripper in pr_review.py to Main #3216.
• Filed Share One Attribute Tail Across the Details and Summary Opener Readers in pr_review.py #3212, End a Code Span at an HTML Block Line in pr_review.py #3214, and Keep an Emphasis Tag Quoted in Code When reduced Strips Emphasis in pr_review.py #3215 from the local review and from Copilot's finding on Strip an Attributed Details or Summary Opener Through TAGS in pr_review.py #3213.
• Answered Copilot's headline on Promote the Attributed Opener Stripper in pr_review.py to Main #3216, which claimed reduced removes code spans before masking, with a probe disproving it.

What not to repeat

• A plain assertion ahead of a subtest loop hides the loop in a revert-proof. On the reverted code, the first plain failure ends the test, so the loop's subtests never run and read as passing. Put plain assertions after the loop, or in a subTest of their own, before trusting which cases fail on the base.
• A tail of "anything to the next >" reads prose as a tag. [^>]*> ate code-span text, and a letter-led [A-Za-z][^>]* still ate <details element's ... ->. Only a real attribute grammar ended it, and HTML_TAG already held one. Search the file for an existing grammar before writing a tail.
• unwrap never masked quotations. Any widening of what TAGS matches reaches text a finding quotes, which is what Copilot caught. A stripper needs the same masking its recognizers already use.
• pr_review.py status exit 50 repeats after the headline is answered. It reads no answer, so that exit on a promotion whose headline caution is answered is not a partial gate.

New learnings

• None beyond the issues filed and the points above.