why i use local llm and ask it to read source code and analysis the code， then it output the json to me ，not execute tool？？？？

[jillsoft-com]

阅读工程源码，分析其功能
read the project and analysis code
{
"name": "grep_files",
"arguments": {
"pattern": ".*"
}
}

为什么我接入本地模型，要求它阅读源码，分析功能时，它给我输出这个json而不是执行tool？？？
why i use local llm and ask it to read source code and analysis the code， then it output the json to me ，not execute tool？？？？

[jillsoft-com]

like this