build-mcp-use-agent
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is a collection of technical documentation, code recipes, and utility scripts designed to assist developers using the mcp-use library. All analyzed content and code serve legitimate development purposes.
- [COMMAND_EXECUTION]: Utility scripts such as
scaffold-agent.shanddiagnose-agent-stuck.shutilize standard shell and Node.js commands (e.g.,npm,node,mkdir,sed) to automate project setup and perform environment diagnostics. These operations are transparent and scoped to the developer's local project directory. - [EXTERNAL_DOWNLOADS]: Documentation and scaffolding scripts refer to well-known, official packages from established registries (npm) and repositories (GitHub), specifically for LangChain LLM providers and official Model Context Protocol (MCP) servers. No downloads from untrusted or suspicious sources were identified.
- [INDIRECT_PROMPT_INJECTION]: The skill identifies the inherent risks of AI agents executing tools based on untrusted data. It provides proactive security guidance, instructing developers to use the
disallowedToolsconfiguration to limit the agent's capability surface area, effectively mitigating potential injection risks through least-privilege principles.
Audit Metadata