The Agent Skills Directory

[SAFE]: The skill establishes a read-only workflow for architectural design, prohibiting any modifications to existing source files and only permitting the creation of planning documents in a specific directory.
[SAFE]: External data retrieval via web_search.py and context7_api.py uses well-known services (DuckDuckGo and Context7) and processes the results as non-executable text, minimizing the risk of remote code execution.
[SAFE]: The init_plan.py script contains robust path validation to prevent path traversal, ensuring that generated plan files remain within the authorized plans/ folder.
[SAFE]: Credential management is handled securely through environment variables for API keys, and the skill does not contain any hardcoded secrets or sensitive information.

plan-mode