skills-best-practices
Pass
Audited by Gen Agent Trust Hub on Apr 5, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is entirely educational, providing guidelines and templates for developers to create high-quality agent skills. It does not contain any executable malicious code.
- [EXTERNAL_DOWNLOADS]: Includes references to official and trusted documentation from Anthropic (anthropic.com, claude.com), the Agent Skills open standard (agentskills.io), and the official Anthropic skills repository on GitHub.
- [COMMAND_EXECUTION]: Contains documentation and examples of shell command usage and dynamic context injection syntax for the Claude Code environment. These are provided as instructional examples to illustrate platform features and do not execute during the loading of this skill.
- [DATA_EXFILTRATION]: Includes specific security warnings about sensitive file paths (e.g., .ssh, .aws, .env) and external fetches to educate developers on potential risks and how to avoid them in their own skill development.
Audit Metadata