The Agent Skills Directory

[SAFE]: The skill implements a structured workflow for architectural analysis using 'Explain' and 'Critique' modes. It defines specific prompts in the references/ directory for sub-agents (explorers, explainers, and critics). The analysis of these prompts and the main SKILL.md reveals that all tasks are strictly defined as read-only code exploration. No execution of code, network calls, or credential access mechanisms are present. The skill uses standard agent orchestration patterns without introducing external dependencies or obfuscation.

how