What problem does it solve? Open pull requests stall in unclear states: unfinished plan steps, missing reviews, unresolved conversations, red CI, base conflicts, or absent QA evidence. This Skill diagnoses exactly what state one PR is in and drives it to merge-ready by dispatching the right companion skills in the right order, instead of making you guess which workflow to run. ## Core Features & Use Cases - State Diagnosis: Collects ten read-only signals (plan progress, diff scope, review decision, CI, mergeability, labels, QA evidence, claim state) into a single PR State Report. - Ordered Skill Chaining: Matches the diagnosis against a state matrix and invokes companion skills verbatim in order — implementation, merge-readiness, QA, then merge — re-checking cheap signals between steps. - Safe Autonomy Controls: Supports --dry-run, --confirm, --allow-merge (off by default), --force, and --max-iterations, with claim locking, stale-lock recovery, and a bounded CI follow-up window. - Use Case: You return to PR #4321 after the weekend and do not know what is left. Run the autopilot to get a full diagnosis, an executed fix-and-review chain, and a published summary comment with labels, QA verdict, and follow-up issues. ## Quick Start Ask the agent to drive PR 4321 to completion with a dry run first, for example: run om-pr-autopilot on PR 4321 with --dry-run and show me the diagnosis and planned chain.
