Skip to content

chore(runbooks): log merge-guard v4.4.51-batch post-install live-probe — 29/29 PASS on installed v4.4.52#1090

Merged
michael-wojcik merged 2 commits into
mainfrom
chore/post-install-probe-v4.4.52
Jul 2, 2026
Merged

chore(runbooks): log merge-guard v4.4.51-batch post-install live-probe — 29/29 PASS on installed v4.4.52#1090
michael-wojcik merged 2 commits into
mainfrom
chore/post-install-probe-v4.4.52

Conversation

@michael-wojcik

@michael-wojcik michael-wojcik commented Jul 2, 2026

Copy link
Copy Markdown
Collaborator

Summary

Logs the post-install live-probe of the merge-guard against the installed v4.4.52 plugin — the final gate held open since PR #1085 (v4.4.51) merged. 29/29 PASS, zero installed-vs-source divergence, no HALT.

Runbook-log-only: +6 lines, append-only new section in pact-plugin/tests/runbooks/RUNBOOK_RUN_DATES.md. No code changes, no version bump (runbook-log precedent: the #1063/v4.4.50 log PR).

What the probe did

  • Byte-identity gate first: full diff -r of the installed v4.4.52 hooks tree vs source at 60f14796 — byte-identical (excl. __pycache__), covering both hook mains + every shared/ module transitively.
  • Verbatim corpus: both pre-merge harnesses recovered from disk (17-row batch set + 14-row fix-level set), deduplicated on the byte-identical N5/D1082-cured form → 29 distinct probes through the installed merge_guard_pre.py/merge_guard_post.py mains as real subprocesses (hooks.json python3 <hook>.py shape, macOS system python3 3.9.6), per-probe isolated CLAUDE_CONFIG_DIR + HOME, zero ~/.claude contamination.
  • Supersede reconciliation: expected verdicts = final shipped semantics; the batch row's historical N5 preserved-DENY probed as ALLOW per the D1082 cure.
  • Harness-fidelity control: the P1a/P1b mint→authorize pair validated the envelope shape before any REFUSE was interpreted as divergence.
  • Outcomes: 17/17 batch-derived · 2/2 fix-level read-floor non-overlap · 5/5 laundering channels REFUSE end-to-end (all four channels, both close-member shapes) · 5/5 faithful round-trips AUTHORIZE.
  • Targeted merge-guard suite in the worktree: 1841 passed / 0 failed / 0 errors.

Issue closure

This row discharges the post-install gate for #1064, #1077, #1078, #1082, #1083 (gated-closure discipline from PR #1085). The five issues will be closed manually after this PR merges, citing the merged row. Not using auto-close keywords.

Not-certified perimeter (mirrored from the row, review M1): NOT probed — #1087 (multi-close ambiguity laundering, open, pre-existing, tracked); #1079 closed won't-fix 2026-07-01.

Copilot AI review requested due to automatic review settings July 2, 2026 18:03

Copilot AI left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Pull request overview

Adds an append-only runbook log entry documenting a post-install live-probe of the merge-guard batch fixes against the installed v4.4.52 plugin, discharging the gated-closure confirmation for the previously-merged v4.4.51 work.

Changes:

  • Appends a new dated section to RUNBOOK_RUN_DATES.md recording a 29/29 PASS post-install probe.
  • Documents byte-identity confirmation (installed hooks vs origin/main), corpus provenance, and outcome breakdown (ALLOW/DENY/REFUSE/AUTHORIZE expectations).

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

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

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants