Progressive Disclosure
A concept originating in UX design, introduced by Anthropic into agent architectures. It employs a two-phase loading strategy—first reading skill names and descriptions, then loading full content on demand—to address LLM context window capacity limits and the Lost-in-the-Middle problem, thereby improving skill routing accuracy.
Core Facts
Progressive Disclosure is originally a classic principle in user interface design, referring to showing complex information only when the user needs it.
Skills enable on-demand loading of specialized knowledge through a progressive disclosure mechanism, loading a particular type of expertise into context only when the task genuinely requires it.
Agent Skill adopts a lazy-loading mechanism whereby the Agent first reads only the metadata (name and description) of each Skill and loads the full instructions only after matching the user's need, so having multiple Skills coexisting does not increase Token consumption
Timeline (last 90 days)
If the catalog descriptions are not accurate enough, the Agent may fail to realize that a useful tool exists and thus take a detour.
Hierarchical loading introduces an extra decision step: the Agent must first determine which type of information it needs, which itself consumes a round of reasoning.
Practical paths for progressive disclosure include hierarchical tool catalogs, on-demand document retrieval, history compression and summarization, and staged system prompts.
A progressive disclosure strategy can make an Agent run 2.3 times faster.
Skill's progressive loading mechanism has two levels: the first level exposes only name and description to form a minimal skill catalog; the second level reads the full SKILL.md and supporting files only after a hit.
Progressive Disclosure refers to an Agent not loading all the full details of every skill at once, but instead gradually expanding relevant information based on task needs
Progressive Disclosure was originally a concept in the UX design field, referring to an interface only showing more details when the user needs them, to avoid information overload
In agent scenarios, RAG is typically encapsulated as a callable tool, with the model deciding on its own when to trigger retrieval, rather than forcibly prepending the full document corpus to every conversation.
The progressive disclosure strategy splits knowledge into separate Skill documents, normally keeping only a short index in the context and dynamically loading the full content only when needed.
Excellent interface design should follow the principle of Progressive Disclosure, presenting information to users only when they need it
3 more timeline events
All Facts (17)
Progressive Disclosure is originally a classic principle in user interface design, referring to showing complex information only when the user needs it.
75%VerifiedSkills enable on-demand loading of specialized knowledge through a progressive disclosure mechanism, loading a particular type of expertise into context only when the task genuinely requires it.
75%VerifiedAgent Skill adopts a lazy-loading mechanism whereby the Agent first reads only the metadata (name and description) of each Skill and loads the full instructions only after matching the user's need, so having multiple Skills coexisting does not increase Token consumption
70%VerifiedSkill's progressive loading mechanism has two levels: the first level exposes only name and description to form a minimal skill catalog; the second level reads the full SKILL.md and supporting files only after a hit.
65%VerifiedProgressive Disclosure was originally a concept in the UX design field, referring to an interface only showing more details when the user needs them, to avoid information overload
65%VerifiedThe core design philosophy of Agent Skills is Progressive Disclosure, meaning that only the information and tools truly needed at the current moment are exposed to the agent at the appropriate time.
65%UnverifiedClaude Code的Skills机制采用渐进式披露（Progressive Disclosure）策略，按需加载专业知识包以防止上下文溢出
85%UnverifiedThe progressive disclosure mechanism means that even if an Agent is equipped with hundreds of skills, the actual context usage during runtime may be equivalent to only that of an MCP solution using 3–5 tools.
60%UnverifiedIf the catalog descriptions are not accurate enough, the Agent may fail to realize that a useful tool exists and thus take a detour.
50%UnverifiedA progressive disclosure strategy can make an Agent run 2.3 times faster.
50%UnverifiedPractical paths for progressive disclosure include hierarchical tool catalogs, on-demand document retrieval, history compression and summarization, and staged system prompts.
50%UnverifiedHierarchical loading introduces an extra decision step: the Agent must first determine which type of information it needs, which itself consumes a round of reasoning.
50%UnverifiedProgressive Disclosure refers to an Agent not loading all the full details of every skill at once, but instead gradually expanding relevant information based on task needs
50%UnverifiedIn agent scenarios, RAG is typically encapsulated as a callable tool, with the model deciding on its own when to trigger retrieval, rather than forcibly prepending the full document corpus to every conversation.
50%UnverifiedThe progressive disclosure strategy splits knowledge into separate Skill documents, normally keeping only a short index in the context and dynamically loading the full content only when needed.
50%UnverifiedExcellent interface design should follow the principle of Progressive Disclosure, presenting information to users only when they need it
50%UnverifiedThe card-based swipe-left/swipe-right interaction follows the Progressive Disclosure principle in cognitive psychology, avoiding decision fatigue associated with long lists.
50%