The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes shell commands (ls, grep) via the Bash tool to manage and search for existing skill documentation within the .claude/skills/ directory.
[SAFE]: The workflow incorporates a 'Quality Gates' phase that explicitly instructs the agent to verify that no sensitive information, such as credentials or personal data, is stored in the generated files.
[NO_CODE]: This skill consists of markdown instructions and YAML configuration for orchestrating agent behavior, without providing external scripts or executables.

learn