truefoundry-applications

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

• Third-party content exposure detected (high risk: 1.00). This skill explicitly fetches and ingests external, user-provided URLs at runtime—e.g., agent_card_url in the "Agent" section and remote OpenAPI spec URLs in "MCP Server (OpenAPI)" (references/manifest-schema.md / references/api-endpoints.md), which are converted into MCP tools or agent behavior and therefore can enable indirect prompt injection.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly allows fetching remote OpenAPI specs at runtime (e.g., https://api.weather.example.com/openapi.json) which are converted into MCP tools that control agent capabilities (and also fetches hosted A2A agent cards like https://research-agent.example.com/.well-known/agent.json), so external URLs can directly control prompts/behavior at runtime.