doc-coauthoring

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

• Third-party content exposure detected (medium risk: 0.30). Stage 1 explicitly instructs the agent to read user-provided links/files and, when integrations are available, to pull in content from monitored sources like shared documents and team channels/threads, meaning outsider-authored free text can be ingested at runtime without selecting a specific item first.