The Agent Skills Directory

[SAFE]: The skill serves as a guide and framework for the creation of other AI agent skills, promoting structured development and high-signal source collection.
[COMMAND_EXECUTION]: Includes a local Python script scripts/quick_validate.py used for structural and quality validation of skill files. The script uses standard libraries and performs safe operations such as regex-based path checking and YAML parsing.
[EXTERNAL_DOWNLOADS]: References multiple external documentation sources including official guides from Anthropic, OpenAI, and Hugging Face, as well as documentation frameworks like Diataxis and DITA. All referenced domains are well-known and reputable industry sources.
[SAFE]: Explicitly instructs users and the agent to avoid storing secrets, credentials, or sensitive data in skill artifacts. The validation script includes automated checks to detect and warn against the use of machine-specific absolute filesystem paths.
[SAFE]: Provides security guidance regarding the use of platform-specific features such as Claude Code hooks and dynamic context injection, warning that these commands run with user permissions and require input sanitization.

skill-writer