Unverified50% confidenceFactExact time
Anthropic's progressive disclosure mechanism works as follows: at startup, only the name and description of each skill (YAML front matter) are read; when the prompt matches a skill, the full skill MD file is read; larger reference files and scripts are loaded on demand.
1
Sources
50%
Confidence
Long-term
Relevance
9/25/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedAnthropic Skills solidifies prompt engineering best practices into reusable files, making them a knowledge asset that can be version-controlled and shared across teams.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75% similarUnverifiedAnthropic Skills adopts a progressive loading mechanism with three layers: the first layer reads only the skill name and description (approximately 100 tokens), the second layer loads the skill.md body, and the third layer invokes specific scripts and reference documents.75% similarUnverifiedThe Skills paradigm injects structured domain expertise into the Agent at startup via a CLAUDE.md configuration file or system prompt.74% similarUnverifiedThe core file structure of Anthropic Skills is a folder containing a skill.md file, which is divided into two parts: YAML metadata (name and description fields) and a Markdown body.73% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/949908API
curl https://kongchang.com/api/v1/knowledge/claims/949908MCP
get_claim(id=949908)