address-github-comments

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 reads GitHub PR comments (user-generated, untrusted content) using "gh pr view --comments" in sub-skills/1-inspect-comments.md, and those comments are used to decide fixes and responses, so they could indirectly inject instructions affecting agent actions.