ingest-video

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

• Third-party content exposure detected (high risk: 0.70). SKILL.md directs the agent to call MCP tools that ingest a user-supplied URL (public web / Google Drive) and then report results (including failure message/error verbatim), meaning outsider-authored text from the fetched page/file metadata or tool error strings can be injected into the agent’s LLM context at runtime via the tool outputs.