The Agent Skills Directory

[SAFE]: The skill is a documentation resource for building agents with the Claude SDK. It contains instructional Python code snippets and architectural patterns.
[COMMAND_EXECUTION]: While the documentation mentions the capability for agents to use a bash tool, it describes this in the context of standard SDK functionality. The examples include security best practices such as input sanitization to block dangerous commands like fork bombs (e.g., :(){ :|:& };:).
[EXTERNAL_DOWNLOADS]: The skill references official and trusted documentation links (docs.claude.com, docs.anthropic.com, modelcontextprotocol.io) and official GitHub repositories from Anthropics. These are well-known, trusted sources for this technology.
[DYNAMIC_CONTEXT_INJECTION]: No !command syntax or load-time execution patterns were detected in the SKILL.md file.
[OBFUSCATION]: No obfuscated URLs, encoded payloads, or hidden characters were found. All URLs and code examples are in plain text.

claude-sdk