AI agents and MCP

Stage 2 · Lesson 3 of 7

Writing the loop in code

Tool calling from scratch

Now write the loop for real with your provider's official in {choice}. Keep a list of messages. Call the model with the messages and tools.

Append the reply. If it asked for tools, run each one, append all the results in one message and loop. Otherwise return the text.

Give two or three simple tools real behaviour, such as reading a file of orders, so results are not faked. Print every call and result while you work, because watching the loop is how you learn it. Add a maximum number of steps from the start.

This small program, perhaps fifty lines, is the core that every framework in stage six wraps.

Common trapjumping to a framework before this works, and then being unable to tell whether a bug is yours or the framework's.

A question to check yourself, an AI practice session and 2 cheat sheet entries come with the full roadmap. Unlock

Resources for this stage

Start with· follow one of these

  • GuideFree
    Claude Tool Use Documentation

    Anthropic · Intermediate

    Read Overview, Define tools and Handle tool calls, then follow the tutorial that grows one call into a full loop

  • GuideFree
    OpenAI Function Calling Guide

    OpenAI · Intermediate

    The same loop on the Responses API: read it after your first provider to see what is universal

Practise

  • Free notebooks: work through the tool use course, which builds tools and a loop exercise by exercise

  • Runnable function-calling examples to compare with your own loop

Keep handy

Also worth your time

Mark this stage as not done?