pr-review

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 1.00). The skill explicitly fetches and ingests user-generated PR content from GitHub via the commands in SKILL.md (e.g., "gh api repos/.../pulls/{N}/reviews", "pulls/{N}/comments", "issues/{N}/comments") and uses those comments to decide fixes, reactions, and replies, so untrusted third-party comments can influence agent actions.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill performs runtime GitHub API calls (e.g., gh api repos/Significant-Gravitas/AutoGPT/pulls/{N}/comments and repos/Significant-Gravitas/AutoGPT/pulls/{N}/reviews) to fetch reviewer comments and then reads and follows those comments to decide fixes and replies, so remote PR comment content directly controls agent instructions at runtime.