skill-creator

Build AI skills using modular components like scripts and references to manage context window usage efficiently. Progressive disclosure defers loading detailed information, keeping core instructions concise while relying on bundled resources for deeper context.

Organize bundled resources by placing executable code in scripts, background documentation in references, and static files in assets. Modular separation ensures the AI agent retrieves precise context without overflowing the context window.

Progressive disclosure in AI skill development is a design pattern that minimizes initial context window consumption. It provides high-level instructions first, then dynamically loads detailed references or scripts only when the task requires deeper specificity.

Create a new AI skill by defining core instructions in the skill file, organizing API documentation in the references directory, and adding helper scripts. This structure ensures efficient resource management and targeted context delivery for the agent.

Modularity and conciseness prevent context window exhaustion and improve agent execution accuracy. Keeping main instructions brief and offloading detailed logic to scripts and references ensures the skill operates efficiently across different tasks.