Teaching an AI about Hands-On AI

You can get an AI assistant (Claude, ChatGPT, and others) to write correct Hands-On AI code for you. There are two easy ways.

Option 1: Paste LLM.txt into the chat

LLM.txt is a single-file reference for the whole package: the four modules, the real function signatures, and worked examples. Drop it into the assistant's context window and it can answer questions and generate working code that uses the current API instead of guessing from older training data.

1. Open LLM.txt (view or raw).
2. Copy its contents, or upload the file, into your AI chat.
3. Ask away. For example:
4. "Using hands-on-ai, write a pirate chatbot."
5. "Build a document Q&A script with the RAG module."
6. "Make an agent with a calculator tool."
7. "Show me a two-stage workflow that researches a topic, then drafts a summary."

Tip: LLM.txt is kept in sync with the package by hand, and it lists the version it describes at the top. If you are on a different version, grab the LLM.txt that matches your install.

Option 2: Chat with the repo on DeepWiki

DeepWiki builds an AI-powered wiki and chat interface over the whole repository. Instead of browsing the code on GitHub, you can ask DeepWiki how something works and it answers with references back to the source.

It is a good way to explore the internals (for example, "how does the agent loop parse tool calls?") without reading every file yourself.

Which one to use

• Writing code with the package? Paste LLM.txt into your assistant (Option 1).
• Understanding how the package works inside? Ask DeepWiki (Option 2).

Both are read-only helpers for learning and building. The package itself does not send your code anywhere.