The Agent Skills Directory

[SAFE]: The skill's primary purpose is to establish security guardrails for AI agents. It provides explicit instructions in Step 3 and the reference documentation to avoid high-risk behaviors like unrestricted command execution, blanket sudo approval, or checking secrets into repositories.
[EXTERNAL_DOWNLOADS]: The skill references official documentation from well-known services including Anthropic (Claude), Google (Gemini), and OpenAI. It also links to the Agent Skills community specification. These references are documented for informational purposes and do not involve dynamic code execution or unverified third-party scripts.

agent-configuration