review

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

• Third-party content exposure detected (medium risk: 0.30). The skill’s /review Greptile triage step fetches outsider-authored PR comment bodies from GitHub (gh api .../pulls/$PR_NUMBER/comments and /issues/$PR_NUMBER/comments) and reads them for classification and reply handling at runtime, so attacker-submitted comment text can be ingested without selecting a specific trusted item.