feedback-builder

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

• Third-party content exposure detected (medium risk: 0.30). The workflow reads bug and feature reports from a dynamic external /feedback HTTP endpoint supplied by users or configuration files.

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

• Potentially malicious external URL detected (medium risk: 0.30). The skill instructs the agent to execute ./secretcurl against an arbitrary remote feedback URL defined in external configuration or user variables without a strong first-party publisher or immutable verification.