The Agent Skills Directory

[SAFE]: The skill performs local reconnaissance using pwd and ls -la to determine the project root and existing documentation. These are standard, low-risk commands for an AI agent to establish context.
[SAFE]: The skill reads project configuration files (e.g., package.json) and explores the codebase using Glob and Grep to understand the application architecture. This is consistent with its stated purpose.
[SAFE]: File writing is restricted to the .claude/rules/ directory for the purpose of storing design documents and project status guidelines. This is a standard practice for persistent agent instructions and does not affect system-level persistence.
[SAFE]: No external network communication, hardcoded credentials, or obfuscated code patterns were detected.

create-app-design