The Agent Skills Directory

[SAFE]: The skill and its associated reference files function as a technical manual for skill development. A thorough analysis confirms that the content is strictly instructional and does not contain any malicious instructions or security vulnerabilities.\n- [COMMAND_EXECUTION]: The documentation provides educational examples of command execution patterns and platform-specific features.\n
Outlines recommended patterns for tool invocation such as uvx, npx, and local script execution in references/format-spec.md.\n
Mentions the shell substitution syntax (!command) available in specific agent platforms in references/platform-guide.md.\n- [EXTERNAL_DOWNLOADS]: The guide includes illustrative examples of how to install skills and extensions across different platforms.\n
References installation commands like pi install and gemini extensions install using placeholder URLs and repository paths in references/platform-guide.md.\n- [PROMPT_INJECTION]: The skill's instructions focus on technical formatting, size optimization, and performance. No patterns indicative of prompt injection, behavior overrides, or security filter bypasses were found.

skill-authoring