Unverified50% confidenceFactExact time
Skills机制与LangChain的Tool Calling、OpenAI的Function Calling同属一脉，核心思想是将LLM推理能力与特定领域工具解耦
1
Sources
50%
Confidence
Long-term
Relevance
7/16/2026
First Seen
Sources
Related Claims
UnverifiedClaude Code中的Skill机制类似于LangChain中的Tool定义或OpenAI的Function Calling声明，本质是提示词的高级封装79% similarUnverifiedOpenClaw的Skill机制本质上是Agent能力的模块化封装，每个Skill对应一个原子级业务能力，拥有JSON Schema定义的输入输出契约77% similarUnverifiedSkill机制的底层是大模型的Function Calling能力，开发者以JSON Schema形式描述工具，模型判断是否调用并输出结构化调用指令77% similarUnverifiedOpenAI将Skill称为Function Calling或Tools，Anthropic的Claude使用Tool Use，微软Copilot生态中对应Plugin与Connector体系76% similarVerifiedSkill是将多步骤的交互逻辑、上下文管理、输出约束打包成一个可调用单元的概念，区别于单次交互的提示词模板76% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/530732API
curl https://kongchang.com/api/v1/knowledge/claims/530732MCP
get_claim(id=530732)