The Agent Skills Directory

[SAFE]: The skill instructions explicitly advocate for security best practices, including using environment variables for secrets instead of hardcoding them and requiring human approval for actions with external side effects.
[COMMAND_EXECUTION]: Includes a Python validation script (scripts/check-output.py) that performs basic file content checks. It is used for verifying deliverable structure and contains no malicious logic or unsafe external interactions.
[EXTERNAL_DOWNLOADS]: References an external MCP endpoint from Plain, a well-known customer service platform. This reference is documented neutrally as part of the implementation guidance.

mcp-setup