create-skill

Progressive disclosure in AI skill development is a best practice that structures documentation hierarchically, providing essential information first while bundling detailed references to optimize agent context usage. This ensures modular and efficient capabilities.

To structure an AI skill for API interaction, you create a SKILL.md file alongside bundled resources like scripts and references. Following a step-by-step creation process ensures the skill is modular, discoverable, and easy for AI agents to use.

Best practices for writing AI skill documentation include maintaining conciseness, applying progressive disclosure, and defining appropriate degrees of freedom. This structure ensures new AI capabilities are well-organized and efficient for agent processing.

You need bundled scripts and references to create effective AI skills when supporting files are required for the task. Organizing these assets alongside SKILL.md ensures the AI agent has direct access to necessary modular components.

The step-by-step process for developing AI skills moves from understanding the skill's core task to structuring content, applying best practices, and iterating. This ensures new capabilities are well-structured and easy for AI agents to use.

You should not use a modular structure for AI skills if it obscures the core task or overcomplicates the documentation. Maintaining conciseness and appropriate degrees of freedom prevents inefficient agent context usage during execution.