url-reader

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

• Third-party content exposure detected (low risk: 0.10). The skill reads arbitrary public web pages via defuddle.md based on URLs specified or requested by the user, representing low-risk indirect prompt injection from selected URLs.

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

• Potentially malicious external URL detected (high risk: 0.70). The skill sends user-provided target URLs to an external third-party public extraction service (defuddle.md) which processes and returns webpage content used to answer the user request, representing an unverifiable third-party operational dependency.