Cover illustration

Architecture diagram

Who this is for

  • Developers who want the model to connect business functions
  • Readers who are learning the basics of Agent

What problem does this article solve?

Models can only generate text and cannot reliably perform queries, calculations, or business actions.

Key concepts

  • tool definition
  • Parameters schema
  • function route
  • Result backfill
  • mock data

Walkthrough

  1. Define callable tools
  2. Let the model decide whether to call
  3. Execute a local function
  4. Give the tool results to the model to generate the final answer

Python code example

The companion demo is generated in:

cd examples/llm-guides/013_tool_calling_weather
python demo.py

The example follows two principles:

  • From OPENAI_API_KEY , load the key at runtime to avoid hard-coding it.
  • Without an API key, the example runs in mock mode so you can follow the workflow.

The essential call looks like this:

import os

api_key = os.getenv("OPENAI_API_KEY")
if not api_key:
    print("未配置 OPENAI_API_KEY,进入 mock 模式。")
else:
    # 真实项目中在这里初始化 SDK 客户端并发起请求。
    print("已读取 OPENAI_API_KEY,可以调用模型。")

Common pitfalls

  • Publishing after a single successful demo run leaves out error handling, logging and human review.
  • Including API keys, real file paths, account details or business data in published articles or code.
  • Readers can struggle to reproduce an example if it is unclear which parts are mocked and which run in production.
  • Citing an API or framework feature without checking its documentation can mislead readers when versions change.

Summary

  • LLM Guide 13 uses function calling to connect a model to local Python functions. A model that can only generate text cannot reliably run a query, perform a calculation, or carry out a business operation itself.
  • Make the workflow reproducible, reversible and easy to review. One good generation tells you little about how it will hold up in use.
  • Before publishing, check current documentation for model capabilities, API parameters and third-party frameworks.

CSDN tags

Function Calling, Tool Calling, Large Model, Python, Agent

WeChat article summary

A local function call illustrates the step from describing an action in words to carrying it out with a tool.

Alternative title for Zhihu

  • What is the problem of Function Calling?
  • How to make a mock call its own Python function?
  • What is the difference between Tool Calling and normal Prompt?

References and further reading

  • Check the OpenAI documentation for your model, SDK and account permissions.
  • Python documentation on environment variables, virtual environments and exception handling.
  • Consult the documentation for the tools used in the example, such as LangChain, LlamaIndex, Haystack or the vector database.