Ask-Samin

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

• Third-party content exposure detected (low risk: 0.10). The workflow requires the agent to search and fetch content from a remote course platform via MCP tools, which constitutes user-selected or searched course content rather than an unmonitored external feed.

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

• Potentially malicious external URL detected (high risk: 0.70). The skill instructs the agent to register and connect an external Model Context Protocol (MCP) server at https://ask-samin-ochre.vercel.app/mcp, which acts as an operational dependency providing runtime tools (search, fetch, browse_curriculum, list_recent_updates) required for the agent's behavior; the publisher provenance of this third-party Vercel app is weak/unknown.