address-pr-comments

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

• Third-party content exposure detected (high risk: 0.95). The skill runs scripts/fetch_comments.py, which uses gh api graphql to fetch PR conversation comments, review bodies, and inline review thread comment bodies (body fields) authored by other GitHub users, then prints them as JSON that becomes LLM-readable context for subsequent steps.

MEDIUM W013: Attempt to modify system services in skill instructions.

• Attempt to modify system services in skill instructions detected (high risk: 1.00). The prompt explicitly instructs running gh commands with "elevated network access" and to rerun with "sandbox_permissions=require_escalated" if sandboxing blocks gh auth status, which requests bypassing sandbox/security mechanisms.