ml-repos

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly documents runtime fetching of external agent cards and OpenAPI specs (e.g. agent_card_url "https://research-agent.example.com/.well-known/agent.json" and remote spec "https://api.weather.example.com/openapi.json"), which are fetched at runtime and converted into MCP tools/agent behavior—allowing remote content to directly control agent capabilities and prompts.