narrative-tracker

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

• Third-party content exposure detected (medium risk: 0.30). SKILL.md’s required workflow ingests outsider-authored free text by running an X.AI x_search query (via curl to https://api.x.ai/v1/responses with tools [{type:"x_search"}]) and then parsing returned output_text content as the primary narrative signal.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill performs a runtime curl POST to "https://api.x.ai/v1/responses" (using an injected XAI_API_KEY) and then parses that API response as the primary narrative signal injected into the agent's processing, so this external URL is used at runtime to supply content that directly controls the agent's output.