sentry

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

• Third-party content exposure detected (high risk: 0.75). The workflow uses the Sentry Web API/CLI to fetch issue/event details (including stack traces, titles, and contexts) from Sentry at runtime; those fields are outsider-authored free text originating from application error reports and user-generated event data, which the agent then ingests into the LLM context.