gh-address-comments

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

• Third-party content exposure detected (high risk: 0.85). The required workflow explicitly fetches GitHub PR review thread comments (outsider-authored text) via the plugin script scripts/fetch_comments.py or an equivalent reviewThreads GraphQL query, and that fetched comment/body text is then used to draft replies—so the LLM context can include outsider free text at runtime.