write-tweet

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

• Third-party content exposure detected (medium risk: 0.30). The runtime workflow’s REMIX and DRAFTS branches fetch outsider-authored free text from X via the X.AI x_search tool using the operator-controlled prompt/topic (e.g., curl to https://api.x.ai/v1/responses with tools:[{type:"x_search"}]), then parse and ingest the returned tweet texts for generation.