find-key-internal-sources

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

• Third-party content exposure detected (high risk: 0.75). The required workflow can read from outsider-authored free text in “Internal Messaging” (public threads/messages and channel topics/purposes) and “Knowledge & Files” (wiki/page body text) at runtime, which the agent would ingest as LLM-readable content for ranking and evidence.