Unverified50% confidenceFactExact time
Skill uses a layered loading design: at startup only the names and descriptions of all Skills are loaded; after a task is issued, descriptions are matched and the full content is then loaded, with on-demand reading during execution.
1
Sources
50%
Confidence
Long-term
Relevance
10/21/2026
First Seen
Sources
Related Entities
Related Claims
VerifiedThe Skill mechanism uses on-demand loading: it first sends the metadata of all Skills to the large model, and after the large model selects the matching Skill, the full instructions are loaded.78% similarVerifiedSkill Creator automatically performs a gap analysis when creating skills, scanning the existing .claude/skills/ directory in the project and comparing new requirements against existing skills to avoid functional overlap or naming conflicts.78% similarUnverifiedHow a Skill is split corresponds to a three-layer structure: large amounts of reference material and expanded content used only in specific situations go into References and are fetched on demand; deterministic operations (sorting, format validation, API calls, calculations) are written as Scripts, and the script code itself does not enter the Context—only the execution results are returned.76% similarVerifiedSkills采用渐进式披露设计，平时只向模型提供技能的名称和描述，具体执行规则在需要时才加载76% similarUnverifiedSKILL.md typically adopts a hierarchical structure: the top layer retains only skill names and functional descriptions, and after a call is determined, detailed instructions, examples, and resource files are dynamically injected76% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1081442API
curl https://kongchang.com/api/v1/knowledge/claims/1081442MCP
get_claim(id=1081442)