docs/use-docs-with-ai.md
Query DataHub docs directly from your AI assistant. We publish a machine-readable index of the documentation so AI coding tools can return accurate, current answers about DataHub — without you switching tabs.
DataHub maintains an llms.txt file: a machine-readable index of the documentation, written for AI tools.
https://docs.datahub.com/llms.txt
Add DataHub docs as a custom source:
https://docs.datahub.com as the URLCursor will index the site. Reference it in chat with @DataHub.
Reference the index in your prompts:
claude "Using https://docs.datahub.com/llms.txt as reference, how do I set up DataHub ingestion from Snowflake?"
Or add the URL to your project's CLAUDE.md so Claude Code uses it on every turn.
Paste the llms.txt URL into the chat:
Use https://docs.datahub.com/llms.txt as reference. How do I write a custom ingestion source in DataHub?
Claude will fetch the index and the relevant linked pages.
With browsing enabled, paste the llms.txt URL into your chat. ChatGPT will use it as a navigation aid for the rest of the conversation.
In VS Code, reference DataHub docs in Copilot Chat using #fetch:
#fetch https://docs.datahub.com/llms.txt explain DataHub's metadata model
Connecting AI to docs is one layer. DataHub also provides AI access to your metadata:
This is an early step toward making DataHub docs first-class for AI workflows. If your AI tool isn't covered above or you have ideas for what to add next, let us know: