bolt-iot-automation

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 call RUBE_SEARCH_TOOLS and related Rube MCP endpoints (e.g., https://rube.app/mcp and composio.dev tool schemas) to fetch tool schemas and recommended execution plans from external third-party servers, which the agent must read and interpret at runtime — exposing it to untrusted third-party content that could carry indirect prompt injections.