fix-pr-comments

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

• Third-party content exposure detected (medium risk: 0.30). The required runtime workflow ingests review thread comment body text from a GitHub pull request via pr-threads.sh (GraphQL reviewThreads → comments → body) and filters only by known bot authors, but the PR author can still contribute free text into those bot threads on fork PRs, creating indirect prompt-injection exposure.