progressive-disclosure

Progressive disclosure optimizes context window usage by loading only the metadata, instructions, or full resources you actually need based on your query and task type. This reduces cognitive load and preserves context for other work.

Progressive disclosure implements a three-tier framework that evaluates query relevance and task complexity to automatically adjust content depth. It upgrades or downgrades the disclosure level as your task evolves, matching the detail to what you need.

Relevance scoring evaluates your query against the skill's content to decide whether to load metadata alone, full instructions, or complete resources. This decision-making adapts to task type and complexity, ensuring you get proportional detail without waste.

Yes. Progressive disclosure supports runtime upgrades—if metadata-only initially loads, you can request deeper guidance and the skill will provide full instructions or resources as your task evolves.

Use progressive disclosure when context is constrained, tasks vary in complexity, or you're exploring a topic before committing to deep investigation. It's particularly valuable for meta-skills and cross-skill loading scenarios where selective content reduces overhead.