tanstack-ai-core-tool-calling

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

• Third-party content exposure detected (medium risk: 0.30). The runtime path ingests outsider-authored free text via the request body’s messages (const { messages } = await request.json()), which the model then uses as prompt input in chat({ ..., messages, tools: [...] }).

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

• Potentially malicious external URL detected (high risk: 0.90). The SKILL connects at runtime to MCP servers (e.g. 'https://mcp.example.com/mcp', 'https://mcp.github.com/mcp', 'https://mcp.linear.app/mcp') via createMCPClient/createMCPClients to discover and invoke remote tools which can execute code or change agent behavior.