Install
$ agentstack add skill-ainova-systems-intelligence-dev-packs-git-finalize-pr ✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
Verified badge
Passed review? Show it. Paste this badge into your README, it links to the public security report.
Reliability & compatibility
Declared compatibility
Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.
We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.
How agent discovery & health will work →About
Finalize the PR
Take a freshly pushed PR through fix/push iterations until CI is green AND every review comment is answered - merge-ready for the owner's accept. Runs autonomously across CI rounds (typically 15-30 min each); wait reactively, never busy-poll.
Pre-flight
git branch --show-current- abort on the default/integration/protected branch.- Resolve the PR:
gh pr list --head --state open --json number --jq '.[0].number'. An explicit argument must match this branch's PR. None - open it withgit-open-prfirst, then retry. - Latest commit:
git rev-parse HEAD. Every check below filters runs byheadSha == HEAD; stale runs for older commits are ignored.
Loop 1 - CI to green
- Probe:
gh pr checksplusgh run list --branch --limit 8 --json databaseId,headSha,name,status,conclusion. - Decide:
- all latest-SHA workflows
success- proceed to Loop 2; - any
in_progress/queued- wait, sized to the longest job's typical duration; - any latest-SHA
failure- drill in.
- Drill in: failed step via
gh run view --json jobs --jq '.jobs[] | select(.conclusion == "failure")'; log viagh run view --job --log(fallbackgh api repos/{owner}/{repo}/actions/jobs//logswhile sibling jobs still run). Find the root-cause file:line; read it and the surrounding code. - Fix smallest-correct, top-down (what happened - what changed since last green - fix or delete per the feature doc - existing primitive?). Run the relevant local gates (
dev-run-tests) before pushing. Batch multi-file fixes into one commit so concurrency cancels prior runs cleanly. - Push, repeat from 1.
Loop 2 - comments drained
- Fetch unresolved threads: GraphQL
reviewThreads(first: 100) { nodes { id isResolved comments(first: 1) { nodes { path line body author { login } } } } }; plusgh pr view --json reviews,comments. - Process via
git-review-pr-comments: fix / discuss / decline-with-reason; reply to every thread. - New pushes restart Loop 1; repeat until green AND drained.
Outcome
Exactly one label, then report (PR URL, label, what needs the owner):
ai:ready-to-merge- green, comments drained,mergeable: MERGEABLE;ai:manual- needs an owner decision (state precisely which);ai:failed- cannot reach green (state the blocking failure and what was tried).
Verify
gh pr checksall green on HEAD; zero unresolved threads; exactly oneai:*label on the PR.
Scope / hand-off
- Merging -
git-merge-prafter the owner accepts. mergeable: CONFLICTING-git-resolve-conflicts, then back into Loop 1.
CRITICAL
- Fix the cause, never the gate: no skipped tests, no loosened checks, no retry-until-green on flaky failures - flag flakes explicitly.
- A failure that already exists on the target branch is reported as pre-existing, never "fixed" on this PR.
- Never merge from this skill, even when everything is green.
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: ainova-systems
- Source: ainova-systems/intelligence-dev-packs
- License: MIT
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet, be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.