address-pr-comments-omo

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

• Third-party content exposure detected (high risk: 1.00). The skill explicitly fetches PR review comments via gh api /repos/{owner}/{repo}/pulls/{pr_number}/comments and includes the untrusted {comment_body} (and diff_hunk) in prompts sent to the hephaestus subagent and used to decide fixes, so arbitrary user-generated review text from GitHub can drive tool actions and thus enable indirect prompt injection.