The Agent Skills Directory

[SAFE]: The skill follows security best practices and provides a professional methodology for extracting documentation from existing codebases without introducing malicious behaviors.\n- [DATA_EXFILTRATION]: While the skill analyzes configuration files and database schemas, it includes strict, repeated instructions to never output actual secret values, credentials, or keys, identifying only their names and purposes.\n- [PROMPT_INJECTION]: The skill processes untrusted local content such as source code and logs. This risk is mitigated through a structured certainty tagging system and mandatory source attribution for every finding extracted from the codebase.\n- [COMMAND_EXECUTION]: Shell access is strictly scoped to a limited set of tools required for system analysis (e.g., database dumpers, container management, and file reading), which prevents the execution of arbitrary or dangerous system commands.

reverse