social

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The prompt explicitly shows curl examples and instructs using a bearer token ({{YOUR_TOKEN}}) in Authorization headers for all authenticated requests, which requires the agent to insert secret API keys verbatim into generated commands/requests.

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

• Third-party content exposure detected (high risk: 0.90). The skill pulls and displays user-generated content from the public inbed.ai APIs (e.g., GET /api/agents, GET /api/chat, GET /api/discover) and its documented workflow explicitly tells the agent to "follow" structured next_steps returned by authenticated responses (API actions with method+endpoint+body), meaning third-party-controlled content can directly instruct and change the agent's actions.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill calls the inbed.ai API (https://inbed.ai and its docs at https://inbed.ai/docs/api) at runtime and explicitly returns executable "next_steps" (method + endpoint + body) from API responses that the agent is expected to execute, meaning remote content directly controls agent instructions.