Expired60% confidenceFactTime unknown
The LLM in Function Calling does not execute code directly, but generates structured descriptions of calling intent, decoupling decision-making from execution
2
Sources
60%
Confidence
Medium-term (~90 days)
Relevance
7/2/2026
First Seen
Valid until: 9/30/2026(expired)
Sources
Understanding Function Calling and MCP Through Cursor's System Prompt
bilibili荒野芯智观察
Related Claims
VerifiedWhen a large model executes Function Calling, it does not actually perform the call itself—it only outputs an intent declaration, and the actual execution is completed by the host application.79% similarVerifiedThe principle of Function Calling is that, in addition to generating natural language text during inference, the model can also output structured function-call instructions (JSON containing the function name and parameters).74% similarUnverifiedFunction Calling本质上是模型决策、外部执行的两阶段流程，模型本身并不直接运行代码73% similarUnverifiedFunction Calling采用关注点分离设计，语言模型只负责决策，真正的工具执行由应用层代码负责，模型不直接触碰外部系统72% similarVerifiedFunction Calling refers to the model declaring, when generating a response, that a specific function needs to be called with specific parameters, which is then executed by an external system and the result returned to the model.70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/54242API
curl https://kongchang.com/api/v1/knowledge/claims/54242MCP
get_claim(id=54242)