kandy

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

• Third-party content exposure detected (high risk: 0.90). The skill instructs the agent to fetch and proxy arbitrary Kandy API endpoints and run actions (e.g., via "membrane request CONNECTION_ID /path/to/endpoint" and action runs that return conversations, call history, voicemails), which exposes it to untrusted, user-generated third-party content from the Kandy platform that the agent is expected to read and act on.