resolve-pr-parallel

SUSPICIOUS: the GitHub CLI usage and data flow are consistent with a PR-review automation skill, but the scope is high-risk because it authorizes autonomous multi-agent code changes, git push, and public PR replies based on untrusted reviewer content. No clear malware or credential-harvesting behavior is shown, but the action authority is broader than a low-risk helper skill.