---
title: Documentation
description: Connect data, query your lake, build models and dashboards, and automate Supernova through its API.
species: guide
---
# Build with your data

Supernova turns the sources you already use into one queryable lake. Connect a source, query its typed tables, save dependable models, and use those results in dashboards or sends.

```sql
select
  currency,
  sum(amount) / 100.0 as revenue
from titan.stripe.charges
where status = 'succeeded' and not _deleted
group by currency
order by revenue desc
```

## Choose a starting point

- [Connect your first source](/v2/docs/quickstart) and carry one Skycrane example from raw data to a model.
- Browse the [connector catalog](/v2/docs/connectors) to see authentication, table shapes, and source-specific edges.
- Learn the [SQL dialect](/v2/docs/queries/sql) and [typesql](/v2/docs/typesql) annotations accepted by the query editor.
- Build with the [public API](/v2/docs/api) or connect an agent through [MCP](/v2/docs/access/mcp).

## How these docs work

Every page has one canonical markdown source. Add `.md` to a page URL to retrieve that source without navigation or presentation. Agents can load the complete corpus from `/v2/docs/llms-full.txt` or discover individual pages through `/v2/docs/llms.txt`.

Examples use fictional sample companies. Identifiers and limits are written exactly as the product accepts them, so you can search the corpus by an error code, directive, route, table, or column.