receiving-code-review

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

• Third-party content exposure detected (high risk: 0.75). The skill’s workflow is triggered by “receiving code review feedback” and then reads the full feedback text; if that feedback comes from external reviewers (e.g., GitHub PR review comments/issues authored by outsiders), that free-form outsider text is ingested into the agent’s LLM context via the runtime review-comment payload.