The Agent Skills Directory

[SAFE]: The skill uses local Python and Shell scripts to perform its tasks. Scripts like audit_claude_md.py and validate_rule_file.py only use standard library modules (json, re, sys, pathlib) and do not perform any network operations or dynamic code execution.
[SAFE]: The init_structure.sh script is an idempotent utility for creating directory structures and template files. It uses standard mkdir and printf commands to write hardcoded templates to the local file system.
[SAFE]: The instructions in SKILL.md and referenced files are focused on improving the clarity and efficiency of AI agent instructions. They include safety-conscious advice, such as avoiding the inclusion of secrets in instruction files.

claude-md