Skip to content

Session Handoff [review-tooling]: Suppressed Unread Check and One-Line Details Bounding Released as 2.0.1087 and 2.0.1088, Tag-Opener Primitive Next #3207

Closed

Description

@ptr727

Next steps, in priority order

  1. Recognize a Details or Summary Opener Through One Primitive in pr_review.py #3206: recognize a <details> or <summary> opener through one primitive in pr_review.py. Seven sites carry four different tag-name boundary rules, and HEADING's <summary branch has none at all. Done looks like: one primitive every site builds on, and a test pinning <summaryx> and <summary plus a no-break space as no opener at every site, and <summary>, <summary plus a tab and an attribute, and <summary/> as an opener.
  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 twice this round, each time on a promotion. 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

State

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

What not to repeat

  • A test claiming a check reports a reader's miss must cover the over-run form, not only a dropped block. The first Check Suppressed Summaries for a Vetted-but-Unread Section in pr_review.py #3193 test stubbed the reader to drop a block. The local review then showed a block running into its sibling counted that sibling's summary as read. Counting every summary a block holds was the defect.
  • A pattern widened to admit markup must be matched by every reader of the same line. Bound a Section Whose Summary Shares Its Details Line in pr_review.py #3204's first cut let HEADING admit a <details> opener while heading_of kept it, so a count-first heading lost its count. A body the base reported as unrecognized became a silent undercount.
  • Python's \s is wider than HTML attribute whitespace. Use [ \t\n\f\r], the set DETAILS_OPEN already uses, after a tag name.
  • A status watcher must read the digest's first line. pr_review.py status prints thread lines after it, and a tail -1 watcher read a thread line as a failed check.
  • Copilot's lite "Needs a closer look" headline can be vague. On Promote One-Line Details Section Bounding in pr_review.py to Main #3205 the specific ask, a tag-name boundary, was only in the body's change list, so read the whole body before answering.

New learnings

  • None beyond the issues filed.

Activity

  1. added
    handoffA link in the session handoff chain, one open issue per track
    on Oct 18, 2026
  2. ptr727 commented on Oct 19, 2026

    @ptr727
    OwnerAuthor

    Continued in #3211, round 6 on track review-tooling.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    handoffA link in the session handoff chain, one open issue per track

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions