What problem does it solve? GitHub Copilot does not reliably re-review a pull request after new commits are pushed, and its review-request API returns misleading responses, so teams risk merging code that was never actually reviewed for the current head commit. This Skill enforces a per-head-SHA review gate so a PR is only considered reviewed when Copilot has covered the exact commit being merged. ## Core Features & Use Cases - Head-scoped review waiting: A background script polls the GitHub API until a Copilot review's commit_id matches the PR's current head SHA, with a 900-second budget and machine-readable STATUS output. - Verdict and thread triage: Reads the review's verdict headline, ignores suppressed-comments blocks by default, and routes open threads to the resolve-pr-feedback skill with the Scope Brief intact. - Documented API behavior: Encodes empirically confirmed GitHub API quirks (inert review requests, no GraphQL fallback, fail-closed thread counts) so agents do not rediscover or misread them. - Use Case: After pushing review fixes to a PR, run this skill to re-request Copilot, wait for a review of the new head commit, resolve any open threads, and confirm the merge gate before merging. ## Quick Start Ask the agent to run the copilot-review skill with the PR number and Scope Brief to wait for and settle the Copilot review on the current head commit.