Unverified85% confidenceFactTime unknown
设置tool_choice为auto时，LLM有时会在应该使用工具时选择直接回答，尤其在使用较小模型时这种行为更加频繁
1
Sources
85%
Confidence
Long-term
Relevance
5/31/2026
First Seen
Sources
LangChain Agent Executor构建原理：从零实现ReAct智能体
youtubeJames Briggs
Related Entities
Related Claims
Unverified设置tool_choice为any等同于required，会强制LLM每次都必须调用工具74% similarUnverifiedWhen an LLM model decides to call a tool, it generates an assistant message containing a tool_use block; after the tool finishes executing, the result is appended as a new message in tool_result format71% similarUnverifiedTool Use（函数调用）允许LLM在推理过程中调用外部工具，如读取文件、执行搜索、调用API或运行代码71% similarUnverified与传统基于规则的专家系统不同，LLM能够根据用户反馈的上下文动态调整排查路径68% similarUnverifiedLLM 0.32a0 通过 execute_tool_calls() 和 reply() 方法将工具调用循环内置到库中66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/16124API
curl https://kongchang.com/api/v1/knowledge/claims/16124MCP
get_claim(id=16124)