fix-coderabbit-review

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

• Third-party content exposure detected (high risk: 0.90). This skill's exporter (scripts/pr-review.ts) explicitly fetches GitHub pull request review comments, issue comments, and review threads (user-generated content) via Octokit/GraphQL and writes ai-docs/reviews-pr-<PR_NUMBER> files which SKILL.md then requires the agent to read/triage and enact (fix code, commit, and resolve threads), so untrusted third-party comments could contain instructions that materially influence the agent's actions.