The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The skill explicitly enforces secure credential management. It includes hard rules to never write secrets into files, instead requiring authentication values to be referenced through environment variable expansion (e.g., ${ENV_VAR}) set by the user in their own environment.
[COMMAND_EXECUTION]: The skill performs benign shell operations for environment verification. During the 'smoke test' phase, it uses commands like test -n to ensure required variables are configured and provides the user with export commands for manual setup.
[DATA_EXFILTRATION]: No unauthorized data access or network exfiltration patterns were detected. The skill operates locally within the project scope.
[REMOTE_CODE_EXECUTION]: The skill generates local configuration files based on a static template. It does not download or execute scripts from remote or untrusted sources.

mcp-agent