skill-creator

You structure an AI skill by creating a SKILL.md file and organizing supporting files into scripts, references, and assets directories. This modular design ensures context efficiency by preventing unnecessary token usage during AI development.

The core principles for creating effective AI skills include maintaining conciseness, setting degrees of freedom for the AI, and using progressive disclosure. These design principles guide how the skill manages context and interacts with the model.

The step-by-step process for developing a new AI skill involves understanding the skill's purpose, structuring the SKILL.md file, managing bundled resources like scripts and references, and iterating on the design. This ensures a modular and efficient build.

Use scripts when your AI skill needs to execute actions or logic, and use references for static configuration files or documentation. Proper resource management ensures the skill loads only necessary context, optimizing token efficiency.

Yes, you can use this guide to build a cloud deployment skill. You would structure your SKILL.md to explain deployment logic, organize configuration files in references, and include deployment scripts in the scripts directory.

Your AI skill likely uses too many tokens because it lacks conciseness and proper progressive disclosure. Adhering to core design principles and efficiently managing bundled resources like references and scripts reduces unnecessary context loading.