Skip to content



Sanity documentation bot

Ask it anything and it reads your documentation and your site's code to give a rich, grounded answer.

entry 13 of 15 · bot · updated 2026-10-07





What it does

The Sanity documentation bot answers questions about your product by reading your documentation and your site's code. Ask it anything about how your schema works, where a field comes from, or why a page renders a certain way, and it works through the source material before it answers.

The answers are grounded in your own material.

How it works

When someone asks a question, the bot searches your documentation and code at the same time. It pulls the relevant sections, cross-references them, and builds an answer that cites what it found. If the code and the docs disagree, it surfaces both. It stays current with your codebase because it reads the live source rather than a snapshot.

What you get

A bot your team and your users can ask in plain language, that cites its sources and keeps pace with your codebase as it changes. Editors and developers both use it. Questions that used to mean a full search through the repo take seconds.

How we set it up

We build it in-house, typically as part of a Sanity website development engagement. The setup covers connecting it to your documentation and your codebase, tuning how it searches and ranks what it finds, and giving it a way to stay in sync as your project evolves. Retainer clients get it wired in alongside the broader build.



Common questions

FAQs

What people ask before picking up Sanity documentation bot

What makes its answers different from a general AI assistant?
It reads your documentation and your site's code rather than a generic model's training data, and it cites what it found. Ask how a schema works, where a field comes from, or why a page renders a certain way, and it works through your source material first.
What happens when the docs and the code disagree?
It surfaces both rather than picking one. That disagreement is usually the thing worth knowing.
Does it go stale as the project changes?
No. It reads the live source rather than a snapshot, so it keeps pace with the codebase.
Who ends up using it?
Editors and developers both, and your users if you expose it. Questions that used to mean a full search through the repo take seconds.
How do we get it?
We build it in-house, typically as part of a Sanity website development engagement. Setup covers connecting it to your documentation and codebase, tuning how it searches and ranks results, and keeping it in sync as the project evolves.


Get in touch

Tell us what you're building. We reply within one working day. Jono or someone on the team picks up every message personally.

By sending this you agree to our privacy policy. We only use your details to reply.