send-dms

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

• Third-party content exposure detected (high risk: 0.80). The skill's "Typical flow" explicitly includes fetching user-generated direct messages via GET /x/dm/{userId}/history (and the Security section notes incoming DM text is untrusted), so the agent will read untrusted third‑party content from X as part of its workflow which could materially influence actions and enable indirect prompt injection.