The Agent Skills Directory

[SAFE]: The skill serves as an educational and operational guide for the standard Model Context Protocol. It does not contain malicious instructions, obfuscation, or persistence mechanisms.
[COMMAND_EXECUTION]: The skill utilizes standard utility commands such as jq, find, which, and grep to validate configuration files and check for existing binaries. These operations are restricted to project-level management and troubleshooting.
[EXTERNAL_DOWNLOADS]: The documentation references the installation of MCP servers via standard package managers (npx, bunx, uvx). The examples provided use well-known and official packages from the @modelcontextprotocol organization and Upstash, which are considered trusted sources for this protocol.
[CREDENTIALS_UNSAFE]: The skill proactively addresses credential security by explicitly warning against hardcoding tokens and demonstrating the correct syntax for using environment variables (${VAR_NAME}) in configuration files.

mcp-management