# Start here

> Three ways in, depending on what you do with flurry.

Canonical: https://dvt.dev/docs/

flurry hosts data apps inside your Snowflake account. You build them by talking to the AI you already use, and every app keeps its versions, its rules for who may change data, and a record of every change.

Pick the door that fits you.

## I use apps

Someone on your team already runs flurry. Open the link they shared, find an app in the Library, and use it. To build your own, [connect your AI](https://dvt.dev/docs/connect/), then try [your first app](https://dvt.dev/docs/first-app/).

## I build apps

You use Claude Code, Cortex Code, Cursor or VS Code. [Connect your AI](https://dvt.dev/docs/connect/) with one command or one click, then paste the first prompt from [your first app](https://dvt.dev/docs/first-app/).

## I run flurry

You are the Snowflake admin. [Install flurry in Snowflake](https://dvt.dev/docs/install/), and if your team uses Claude, [let them use it with flurry](https://dvt.dev/docs/claude-connector/). To try it first on a laptop, see [try flurry on your laptop](https://dvt.dev/docs/local/).

## Good to know

- [How flurry works](https://dvt.dev/docs/how-it-works/): versions, rules for writes, and the audit trail.
- [What leaves your account](https://dvt.dev/docs/what-leaves-your-account/): nothing, unless you turn on a connector.
- [Troubleshooting](https://dvt.dev/docs/troubleshooting/): each error and its fix.
- AI agents start at [the agent runbook](https://dvt.dev/docs/agents/).