Purpose
This project provides a collection of reusable skills and agents for Codex and Claude. It includes specialized tools to convert Markdown into Xiaohongshu image cards, process Gmail invoices, generate medical presentations, and add academic citations.

Security Assessment
The overall risk is rated as Medium. While the tool does not request inherently dangerous permissions, the codebase contains a few flagged security points. A recursive force deletion command (`rm -rf`) was found in a package shell script, which could potentially be destructive if misconfigured or executed in the wrong directory. Additionally, a script accesses environment variables, likely for handling API keys or credentials. Users should carefully inspect the code to ensure no sensitive data is exposed and that system directories are protected from unexpected deletion. No hardcoded secrets were detected, though network requests are likely required for AI integrations and email processing.

Quality Assessment
The project appears active and reasonably maintained, with its latest push occurring today. It is backed by a standard MIT license, which is excellent for open-source transparency and legal clarity. The community trust level is still in its early stages, currently sitting at 11 GitHub stars, indicating a small but growing user base.

Verdict
Use with caution. The project is active and properly licensed, but developers should review the shell scripts and environment variable usage before executing anything locally.