The Agent Skills Directory

[SAFE]: The skill consists of documentation, reference guides, and configuration templates for an AI agent's skill system. No malicious code or obfuscated patterns were detected during analysis.
[COMMAND_EXECUTION]: The documentation provides standard CLI commands for testing and configuring hooks, such as using 'npx tsx' to run local TypeScript scripts, 'npm install' for dependency management, and 'chmod +x' for setting file permissions. These are intended for the local development environment.
[PROMPT_INJECTION]: The skill describes the implementation of a 'UserPromptSubmit' hook designed to inject context into the model's prompt. This is an intended architectural feature of the system being documented and does not represent a malicious bypass or override.

skill-developer