The Agent Skills Directory

[SAFE]: The skill provides legitimate guidance on creating MCP servers using established software development kits and platform-specific CLI commands (e.g., copilot mcp).\n- [SAFE]: Encourages security best practices by recommending the centralization of transport concerns such as authentication, header injection, and OAuth token refreshes in a reusable middleware layer.\n- [SAFE]: References standard libraries such as @modelcontextprotocol/sdk and zod for implementation, which are industry-standard tools for this task.\n- [SAFE]: No malicious patterns, obfuscation, or unauthorized data access were detected. All external references are for documentation and optimization examples.

mcp-builder