An Illustrated Guide to AI Agents

Chapter 3. Tool Usage, Learning, and Protocols

A Note for Early Release Readers

With Early Release ebooks, you get books in their earliest form—the author’s raw and unedited content as they write—so you can take advantage of these technologies long before the official release of these titles.

This will be the 5th chapter of the final book. Please note that the GitHub repo will be made active later on.

If you’d like to be actively involved in reviewing and commenting on this draft, please reach out to the editor at [e-mail].

The tool module is such an interesting module for augmenting your LLM and generating autonomous behavior in agents. By themselves, LLMs are nothing more than functions that take in some string and output some string. Although that creates interesting chatbots, they cannot yet interact with their environment.

A key trait that defines an agent is its ability to autonomously search for, select, and utilize tools, allowing it to interact with and influence its environment. With enough capabilities, agents can even create their own set of tools to use. The benefit of tools is not contained to interaction with the environment. Tools are typically used to access external knowledge or memories, as we discussed in the previous chapter. We can even use tools to access specialized (L)LMs that have capabilities that extend beyond what the agent is capable of, like multimodal LLMs or LLMs specialized for certain tasks (like coding). Generally, tools allow an ...