x-undocumented-api

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

• Third-party content exposure detected (medium risk: 0.65). This skill’s runtime workflow can fetch arbitrary outsider-authored web JavaScript/HTML from https://x.com and then ingests extracted free-text operation metadata from those bundles into the agent context via scripts/extract-x-graphql-endpoints.ts (public web content fetch/scrape path).

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

• Potentially malicious external URL detected (high risk: 0.90). The included extraction script and references fetch https://x.com and then load client scripts from https://abs.twimg.com/responsive-web/client-web/*.js at runtime to extract query IDs/operation metadata that the agent uses, so these external URLs directly influence runtime behavior and operation selection.