hunt-grpc

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

• Third-party content exposure detected (low risk: 0.10). The required workflow actively calls user-chosen target endpoints (gRPC reflection, RPC calls, and transcoded routes) and reads their responses, where outsider-authored content such as user profile fields could appear, but the agent must select and fetch each call, so the exposure is low rather than medium.

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

• Potentially malicious external URL detected (medium risk: 0.30). The skill instructs cloning and building the individually owned GitHub repository secengjeff/rapidresetclient during setup and then executing the resulting binary, which is an operational code dependency. The publisher is an individual rather than a recognized vendor or project organization, so it is unofficial, but it shows no concrete warning sign, making it common/reputable unofficial and medium.