om-auto-review-pr

Reviews pull requests in isolated worktrees and runs autonomous autofix loops until merge-ready.

1|Updated Apr 23, 2026
One-click install
npx skills add https://github.com/tkogut/agents-os-core --skill om-auto-review-pr-tkogut
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: om-auto-review-pr
Source: https://github.com/tkogut/agents-os-core/tree/main/global_skills/om-auto-review-pr
Command: npx skills add https://github.com/tkogut/agents-os-core --skill om-auto-review-pr-tkogut

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Reviewing pull requests manually is slow and inconsistent, and automation that waits on CI or clobbers other agents' work creates bottlenecks and conflicts. This Skill reviews or re-reviews any PR by number in an isolated worktree, submits an approve or request-changes verdict immediately, manages pipeline labels, and — when eligible — autonomously fixes findings, resolves conflicts, and re-reviews until the PR is merge-ready. ## Core Features & Use Cases - Isolated PR review: Fetches the exact PR from the tracker, checks it out in a temporary worktree, and runs the full om-code-review engine plus diff-level auto-detections without touching your working tree. - Autonomous autofix loop: On the automation's own PRs (or with --autofix), resolves merge conflicts first, fixes all actionable findings including inherited reviewer feedback, reruns validation, and re-reviews until approvable. - Claim and label discipline: Uses a three-signal in-progress lock, pipeline label transitions, priority/risk inference, and bounded CI follow-up so concurrent agents never clobber each other. - Use Case: A teammate asks you to review PR #1234. Invoke the skill with the PR number; it claims the PR, reviews the diff, posts a changes-requested verdict with findings, hands it back to the author — or, with --autofix, fixes everything and pushes a merge-ready update. ## Quick Start Ask the agent to review pull request number 1234 and automatically fix any findings it can resolve.

Frequently Asked Questions about om-auto-review-pr

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I automatically review a pull request with an AI agent?▼

Invoke the skill with the PR number, for example /om-auto-review-pr 1234. It claims the PR, checks it out in an isolated worktree, runs the full code review, and submits an approve or changes-requested verdict with labels and a handoff comment.

How does the autofix loop work on pull requests?▼

When the PR author is the automation itself or --autofix is passed, the skill resolves merge conflicts first, fixes all actionable review findings including inherited reviewer comments, reruns validation commands, and re-reviews until the PR is approvable or a real blocker remains.

Can it review pull requests from forked repositories?▼

Yes. For fork PRs it preserves the original head SHA and authorship, creates a carry-forward branch in the main repository, applies fixes there, opens a replacement PR crediting the original author, and closes the original only after the replacement exists.

Does the review wait for CI checks to pass before submitting?▼

No. The verdict, labels, and comments are submitted immediately after the review, with pending checks disclosed in the review body. A bounded follow-up watches CI for at most the configured maxWaitMinutes (default 40) and posts the result separately.

What happens if another agent is already reviewing the same PR?▼

A three-signal check (in-progress label, foreign assignee, recent claim comment) detects the existing lock and the run stops, asking the user before overriding. Passing --force takes over with an explicit override comment naming the previous owner.

When should I not use automated PR review?▼

Avoid it when changes involve ambiguous product or architecture decisions, missing credentials or infrastructure access, or scope changes that would break public contracts — the skill stops and asks for human judgment in exactly these situations.