feat(pr-pipeline): let a trusted author's gate-path PR clear on a passed security review by kyle-sexton · Pull Request #10708 · melodic-software/claude-code-plugins

kyle-sexton · 2026-10-28T18:00:47Z

No related issue: owner direction 2026-10-27 that hold-floor PRs auto-merge unless HIGH/CRITICAL

Summary

A gate-class hold-floor path now clears on a passed security review, the way an enforcement path does, but only when all of these hold:

the pull request was opened by a person on the trusted-actor list (a User, never the lanes App or any other bot);
the gate paths it changes were touched only by that person's own commits;
pr-verify passed the head being judged.

trust-class paths, paths a configured denied-path glob also matches, and lanes App pull requests stay person-held. A lane's fix commit on a gate path, or commits that cannot be read, keep the hold too, so a lane changing its own powers is still merged by a person.

Fix

babysit_risk.py: adds AUTHOR_REVIEW_FLOOR (the gate globs) and RiskInput.author_reviewed. A path in that set that no trust or configured glob matches rates review-path: instead of high-path:, so a clean security review bound to the head clears it. Any value other than a frozenset of strings lifts nothing.
babysit_merge_policy.py: author_reviewed_paths computes that set from the PR author, the trusted-actor list's people and foreign_paths, and returns nothing when pr-verify did not pass or the author is not a listed person. assess passes the trusted people through.
babysit_merge.py: score_unattended computes the set only on a complete file listing, with verified meaning the verifier reader ran.
docs/conventions/pr-pipeline/README.md states the new clear condition under the hold-floor classes; changelog fragments added.

Verification

Run on head efb51b4, which merges main 9d2beb6 (#10681, scripts/check-affected.sh only, clean merge) into 981f125 (itself a merge of main f714f1e). Each suite was also run on 981f125 first, with the same counts.

uvx pytest plugins/source-control/skills/babysit-prs/scripts/tests/ -q: exit 0; 3369 passed, 5 skipped, 4109 subtests passed.
python3 plugins/source-control/skills/babysit-prs/scripts/tests/test_merge_holds_sync.py: exit 0; 15 tests, OK.
node --test .github/scripts/pr-merge/*.test.mjs: exit 0; 617 tests, 617 pass, 0 fail.
scripts/validate-plugins.sh --only "source-control": exit 0; plugin contracts validated; 583 mod tests pass, 0 fail, across 8 test files; all plugin manifests and the catalog validated.
Ready verify gate (scripts/check-affected.sh --tests) on efb51b4: pass, 14 passed, 2 skipped.

Left out: no live babysit run against a real trusted-author gate-path PR; the unit tests cover the policy and risk paths.

…sed security review A pull request a person on the trusted-actor list opened now clears a `gate` floor path that only that person's own commits touched, on a head pr-verify passed, once the security review binds to the head with no open finding, as an `enforcement` path clears. Lanes App and bot pull requests, a lane's fix commit on a gate path, unreadable commits or lists, and every `trust` path still wait for a person. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

…for-trusted-author # Conflicts: # docs/conventions/pr-pipeline/README.md # plugins/source-control/skills/babysit-prs/scripts/babysit_risk.py # plugins/source-control/skills/babysit-prs/scripts/tests/test_merge_holds_sync.py

…-hold-clears-for-trusted-author

chatgpt-codex-connector · 2026-10-28T18:09:23Z

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-10-28T18:11:56.436405Z	`efb51b4`	Draft marked ready

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

Open a pull request for review
Mark a draft as ready
Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

github-actions · 2026-10-28T18:12:10Z

claude-review reviewed 6ce56e9d93317dbee4405ab0e5259cfb37ec32e2: 1 finding(s), 0 as inline comments.

Incremental review of 83c4d86..6ce56e9. The change moves enable-auto-merge.mjs, superseded.mjs, clear-escalations.mjs, gate-check.mjs and head-ref-map.mjs from gate to trust. That takes them out of the trusted-author lift, which closes a real gap: an owner-signed PR could otherwise have weakened a hold on a clean security review. The change is correct and hardens the gate. I read the two class tables, the new test and the docs, and found one doc/comment inconsistency. I did not run any tests.

Checked

The two class tables agree. babysit_risk.FLOOR_CLASSES and the merge-holds.mjs HOLD_PATH_CLASSES copy carry the same five reclassifications.
The new counts match the table. The merge-holds.mjs block has 14 gate rows left (13 + 5 = 18 trust, 19 − 5 = 14 gate), enforcement stays at 43, and the total stays 75. HIGH_FLOOR (trust plus gate) stays 32, so test_the_person_floor_is_the_same_list is unaffected.
AUTHOR_REVIEW_FLOOR is derived from the gate class, so the five scripts drop out of it automatically. test_the_author_lift_is_the_gate_class_and_the_lanes_app_never_gets_it still holds. The new test asserts the exclusion and that the path rates high-path: even after a clean review.
test_every_pr_merge_script_is_held_for_a_person accepts any PERSON_CLASSES member, so it passes with trust.
The existing floor tests that list these five paths only check that they are held, not their class, so the reclassification doesn't break them.
The change does not lower any protection and has no cost impact. Nothing in it reaches the critical-security or cost hold classes.

Finding (advisory): two comments still describe the old class split

The HOLD_PATH_CLASSES header comment in .github/scripts/pr-merge/merge-holds.mjs was not updated. Its trust list (L247-L256) stops at "this module, and scripts/sync-hold-floor.py". It defines gate as "the other pr-merge scripts and the rate guard". After this change five pr-merge scripts are trust. The same comment in babysit_risk.py was updated in this commit, so the two copies of the comment now disagree. The comment sits outside the generated BEGIN/END block, so sync-hold-floor.py will not repair it.
The README gate bullet (L939-L942) still says gate covers "every other script of .github/scripts/pr-merge/". Elsewhere the same README now lists the five as trust, so "every other" reads as ambiguous. Saying "every other script of .github/scripts/pr-merge/ that the trust bullet does not list" would fix it.
Neither comment affects behavior. A reader auditing the classes from these comments would believe those five scripts still take the author lift.

No inline comment is attached, because neither location is on a changed line.

github-actions

claude-security-review: 1 inline finding(s) on efb51b407b344f264619c0e5333dc2f92f48b0b7; the summary is in this lane's pull request comment.

claude-lane: claude-security-review

github-actions · 2026-10-28T18:12:36Z

claude-security-review reviewed 6ce56e9d93317dbee4405ab0e5259cfb37ec32e2: 0 finding(s), 0 as inline comments.

No security issues found in the incremental change since 83c4d86.

The change only tightens holds. It reclassifies five hold-applying pr-merge scripts from gate to trust in the floor tables: enable-auto-merge.mjs, superseded.mjs, clear-escalations.mjs, gate-check.mjs and head-ref-map.mjs. A trust path never takes the trusted-author lift, so an owner-signed PR that weakens one of these scripts now waits for a person. Before, it could clear on a passed security review.

What I checked:

Tables agree. babysit_risk.py:261-266 and merge-holds.mjs:278-283 carry the same five trust rows. Each file has 18 trust entries, which matches the test's updated counts (trust 13 to 18, gate 19 to 14, enforcement 43 unchanged).
Nothing widens. The change adds no entry to AUTHOR_REVIEW_FLOOR, and it doesn't change who is trusted, what a token may do, or the kill switch. The new test asserts that all five paths are outside AUTHOR_REVIEW_FLOOR and hold after a clean security review.
No leftover gate assertion in the tests. The test in test_babysit_risk.py that checks the floor holds with the switch off still lists these paths, which is correct. The test_merge_holds_sync.py change is a docstring only.
Docs and changelog fragments match the code.

I did not run the test suites.

critical-findings: 0

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

…-hold-clears-for-trusted-author

The author lift clears the whole `gate` floor class on a trusted author's own pull request after a clean security review. Agents commit as the trusted owner, so a pull request widening a lane token's scope through resolve-config (effect-grants.json, lane-rules.json, activity-rules.json, the broker's signer code) could merge with no person. Reclassify .github/actions/resolve-config/** from `gate` to `trust`: both classes hold for a person, and every other reader (HIGH_FLOOR, merge-holds.mjs HOLD_PATH_FLOOR, hold-paths, hold_patterns) treats them alike, so only the lift changes. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

…for-trusted-author # Conflicts: # plugins/source-control/skills/babysit-prs/scripts/tests/test_babysit_risk.py

…mments Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

…-hold-clears-for-trusted-author

github-actions

claude-security-review: 1 inline finding(s) on cfaaa2d6a05e16ebbee7dfd250d7f46628bbc201; the summary is in this lane's pull request comment.

claude-lane: claude-security-review

github-actions

claude-review: 1 inline finding(s) on cfaaa2d6a05e16ebbee7dfd250d7f46628bbc201; the summary is in this lane's pull request comment.

claude-lane: claude-review

Lane-Activity: pr-unblock/resolve-conflicts

…for-trusted-author # Conflicts: # plugins/source-control/skills/babysit-prs/scripts/tests/test_babysit_risk.py

…s-for-trusted-author' into fix/gate-hold-clears-for-trusted-author

github-actions

claude-security-review: 1 inline finding(s) on fcd3629631601bcf3b3b1f339a8d4a9bfb0a458a; the summary is in this lane's pull request comment.

claude-lane: claude-security-review

melodic-automation-lanes · 2026-10-28T21:41:35Z

Unattended merge: risk HIGH at head 0140ff1cd13cc37eeac02d58eb9bbed83d672313. Only LOW, or MED on class, size and reported review-finding reasons alone, merges without a human. A floor path holds it: a human merges it in person, never an agent or a lane.

HIGH high-path:.github/scripts/pr-merge/merge-holds.mjs: .github/scripts/pr-merge/merge-holds.mjs
HIGH high-path:plugins/source-control/skills/babysit-prs/scripts/babysit_merge.py: plugins/source-control/skills/babysit-prs/scripts/babysit_merge.py
HIGH high-path:plugins/source-control/skills/babysit-prs/scripts/babysit_merge_policy.py: plugins/source-control/skills/babysit-prs/scripts/babysit_merge_policy.py
HIGH high-path:plugins/source-control/skills/babysit-prs/scripts/babysit_risk.py: plugins/source-control/skills/babysit-prs/scripts/babysit_risk.py
MED findings:claude-review: 1 open, reported only
MED findings:claude-security-review: 3 open below CRITICAL, reported only

docs/conventions/pr-pipeline.yaml sets the activity effect values and lane slots the token broker reads, so it scopes lane tokens as resolve-config does: classify it trust, so a trusted author's own pull request no longer clears it on a security review alone. Bring the HOLD_PATH_CLASSES comment in merge-holds.mjs in line with the trust class of resolve-config. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

…-hold-clears-for-trusted-author

github-actions

claude-security-review: 1 inline finding(s) on 11abd85bf15b8094e4c53de9c10d8303e0e27be6; the summary is in this lane's pull request comment.

claude-lane: claude-security-review

… lift An agent signs commits as the trusted owner, so an owner-signed pull request could widen the author lift or drop a floor row on a clean security review, and the next pull request would take the wider lift. Classify the code that computes the lift and the hold tables trust: babysit_risk.py, babysit_review_evidence.py, babysit_merge_policy.py, babysit_merge.py, merge-holds.mjs and scripts/sync-hold-floor.py. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

…-hold-clears-for-trusted-author

Lane-Activity: pr-unblock/resolve-conflicts

gitguardian · 2026-10-29T01:11:38Z

️✅ There are no secrets present in this pull request anymore.

If these secrets were true positive and are still valid, we highly recommend you to revoke them.
While these secrets were previously flagged, we no longer have a reference to the
specific commits where they were detected. Once a secret has been leaked into a git
repository, you should consider it compromised, even if it was deleted immediately.
Find here more information about risks.

^{_{🦉 GitGuardian detects secrets in your source code to help developers and security teams secure the modern development process. You are seeing this because you or someone else with access to this repository has authorized GitGuardian to scan your pull request.}}

github-actions

claude-security-review: 1 inline finding(s) on 83c4d86c6fddbce95acf740dfda8b00af3a78dc0; the summary is in this lane's pull request comment.

claude-lane: claude-security-review

…author lift enable-auto-merge.mjs, superseded.mjs, clear-escalations.mjs, gate-check.mjs and head-ref-map.mjs apply and enforce the holds. Agents commit as the trusted owner, so a gate-class lift on them would let an owner-signed PR weaken a hold on a clean security review, and the weakened script would judge every later PR. They move to trust, as merge-holds.mjs did. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

github-actions · 2026-10-29T02:49:37Z

Post-merge mutation testing

Range: 0c235dd..266d444
Merged pull request: #10708
Outcome: refused

Killed 0, survived 0, timeout 0, no coverage 0, invalid 0, not run 7.
Covered-code score n/a; mutation score n/a.

File	Killed	Survived	Timeout	No coverage	Invalid	Not run
`.github/scripts/pr-merge/merge-holds.mjs`	0	0	0	0	0	1
`plugins/source-control/skills/babysit-prs/scripts/babysit_merge.py`	0	0	0	0	0	1
`plugins/source-control/skills/babysit-prs/scripts/babysit_merge_policy.py`	0	0	0	0	0	2
`plugins/source-control/skills/babysit-prs/scripts/babysit_risk.py`	0	0	0	0	0	3

The run's summary has the full report.

kyle-sexton and others added 4 commits October 28, 2026 13:29

Merge commit 'f714f1e11254d3dbd1bb2ce4fb33ed293f9677b8' into fix/gate…

981f125

…-hold-clears-for-trusted-author

Merge commit '9d2beb633dfd0f6bf2a27dc1dd430f50a5dd1ff0' into fix/gate…

efb51b4

…-hold-clears-for-trusted-author

kyle-sexton marked this pull request as ready for review October 28, 2026 18:09

github-actions Bot reviewed Oct 28, 2026

View reviewed changes

Comment thread plugins/source-control/skills/babysit-prs/scripts/babysit_risk.py Resolved

kyle-sexton and others added 6 commits October 28, 2026 14:15

docs(pr-pipeline): state the owner direction without the quote

c4ea079

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

Merge commit 'ac3e67e6dfb8cc52afd8cafa206ff1fc0743c550' into fix/gate…

beb46d3

…-hold-clears-for-trusted-author

Merge remote-tracking branch 'origin/main' into fix/gate-hold-clears-…

71a53e1

…for-trusted-author # Conflicts: # plugins/source-control/skills/babysit-prs/scripts/tests/test_babysit_risk.py

test(pr-pipeline): drop the PR reference from the token-scope test co…

f3e7dd9

…mments Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

Merge commit 'fe893e18bd372765a400060b148bdc749c4abb59' into fix/gate…

cfaaa2d

…-hold-clears-for-trusted-author

github-actions Bot reviewed Oct 28, 2026

View reviewed changes

Comment thread plugins/source-control/skills/babysit-prs/scripts/babysit_merge_policy.py Resolved

github-actions Bot reviewed Oct 28, 2026

View reviewed changes

Comment thread .github/scripts/pr-merge/merge-holds.mjs Resolved

melodic-automation-lanes Bot and others added 3 commits October 28, 2026 19:20

Merge main into fix/gate-hold-clears-for-trusted-author

fcd3629

Lane-Activity: pr-unblock/resolve-conflicts

Merge remote-tracking branch 'origin/main' into fix/gate-hold-clears-…

cf22662

…for-trusted-author # Conflicts: # plugins/source-control/skills/babysit-prs/scripts/tests/test_babysit_risk.py

Merge remote-tracking branch 'refs/remotes/origin/fix/gate-hold-clear…

0140ff1

…s-for-trusted-author' into fix/gate-hold-clears-for-trusted-author

github-actions Bot reviewed Oct 28, 2026

View reviewed changes

Comment thread plugins/source-control/skills/babysit-prs/scripts/babysit_risk.py Resolved

kyle-sexton mentioned this pull request Oct 28, 2026

feat(pr-pipeline): bind workflow-edit reviews to a run from main #10743

Merged

kyle-sexton and others added 2 commits October 28, 2026 19:00

Merge commit '91d7250d4d413a019a59813a4cbc67c64a8c2934' into fix/gate…

11abd85

…-hold-clears-for-trusted-author

github-actions Bot reviewed Oct 28, 2026

View reviewed changes

Comment thread plugins/source-control/skills/babysit-prs/scripts/babysit_merge_policy.py Resolved

kyle-sexton and others added 2 commits October 28, 2026 19:10

Merge commit '4b544da30e39fafa1cd8f5972d81a7ac476409d3' into fix/gate…

cee46c6

…-hold-clears-for-trusted-author

kyle-sexton mentioned this pull request Oct 29, 2026

fix(pr-pipeline): align every needs-human path with the owner stop rule #10517

Closed

14 tasks

kyle-sexton added this pull request to the merge queue Oct 29, 2026

github-merge-queue Bot removed this pull request from the merge queue due to a conflict with the base branch Oct 29, 2026

Merge main into fix/gate-hold-clears-for-trusted-author

83c4d86

Lane-Activity: pr-unblock/resolve-conflicts

github-actions Bot reviewed Oct 29, 2026

View reviewed changes

Comment thread plugins/source-control/skills/babysit-prs/scripts/babysit_risk.py Resolved

kyle-sexton added this pull request to the merge queue Oct 29, 2026

Merged via the queue into main with commit 266d444 Oct 29, 2026
41 checks passed

kyle-sexton deleted the fix/gate-hold-clears-for-trusted-author branch October 29, 2026 02:11

This was referenced Oct 29, 2026

feat(source-control): resolve lane-answered outdated threads on owner PRs in the worker tier #10841

Merged

fix(source-control): close the cost hold's billed and security probe gaps #10583

Merged

kyle-sexton mentioned this pull request Oct 29, 2026

fix(source-control): stop holding non-billing model and timeout changes as cost #10866

Merged

Conversation

kyle-sexton commented Oct 28, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Fix

Verification

Related

Uh oh!

chatgpt-codex-connector Bot commented Oct 28, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codex Review Summary

Uh oh!

github-actions Bot commented Oct 28, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

github-actions Bot left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

github-actions Bot commented Oct 28, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

github-actions Bot left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

github-actions Bot left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

github-actions Bot left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

melodic-automation-lanes Bot commented Oct 28, 2026

Uh oh!

github-actions Bot left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

gitguardian Bot commented Oct 29, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

️✅ There are no secrets present in this pull request anymore.

Uh oh!

github-actions Bot left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

github-actions Bot commented Oct 29, 2026

Post-merge mutation testing

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

kyle-sexton commented Oct 28, 2026 •

edited

Loading

chatgpt-codex-connector Bot commented Oct 28, 2026 •

edited

Loading

github-actions Bot commented Oct 28, 2026 •

edited

Loading

github-actions Bot commented Oct 28, 2026 •

edited

Loading

gitguardian Bot commented Oct 29, 2026 •

edited

Loading