xquik-twitter

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

• Third-party content exposure detected (high risk: 0.85). Outsider-authored free text from X (e.g., tweet/article/body text, bios, DM text, and error messages) is ingested into the agent’s LLM context via runtime API/MCP reads like GET /api/v1/x/tweets/{id}, GET /api/v1/x/articles/{tweetId}, GET /api/v1/x/users/{id}, GET /api/v1/x/dm/{userId}/history, and GET /api/v1/events (event data.text), which the workflow explicitly treats as untrusted but still passes as readable text to the model.