mercury-mcp

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The skill includes an explicit example of embedding an API key in an HTTP header / CLI command (-H "x-api-key: ak_agent_..."), which encourages placing secret values verbatim into generated commands or outputs.

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

• Potentially malicious external URL detected (high risk: 1.00). The skill explicitly uses the external JSON-RPC endpoint https://api.mercury.build/api/v1/mcp at runtime, which can return agent configuration including system prompts (e.g., via mercury_get_agent_context and admin_get_agent_details) that directly control the agent's instructions.