Connector directory

SQL Server. Stripe. A folder of CSVs. Same SQL.

Relational engines, hosted cloud databases, business APIs and folders of files — all reachable with the same read-only SQL through one outbound agent. Open any card for how it works.

Try the Free Tier Team Pricing

Free forever tier · No credit card required

New · Files & folders

A folder of files is a database now.

Point the agent at a drop-folder of CSVs, Parquet, Excel, SQLite or JSONL. First sync pins the schema. Query the whole set, or one file, with the same read-only SQL you already use on Postgres.

NewCSV FolderSpreadsheet filesSetup guideNewExcel FolderWorkbooksSetup guideNewFile SystemFolder metadataSetup guideNewJSONL FolderEvent logsSetup guideNewMulti-record CSVTagged recordsSetup guideNewParquet FolderColumnar filesSetup guideNewSQLite FolderEmbedded databasesSetup guide

The newest seven of 53 connectors. See every data source

folder_revenue.sqlSQL
-- One folder of CSVs. Whole set, or one file.
SELECT   _source_file, SUM(qty) AS units, SUM(price) AS revenue
FROM     orders
GROUP BY _source_file
ORDER BY revenue DESC
All 53 connectors

Find your data source

Everything we connect to, in one place — databases, hosted services, business APIs and folders of files. Type a name or filter by kind. Every card opens a short brief: what it is, how it works, where to start.

Showing 53 of 53

NewCSV FolderSpreadsheet filesSetup guideNewExcel FolderWorkbooksSetup guideNewFile SystemFolder metadataSetup guideNewJSONL FolderEvent logsSetup guideNewMulti-record CSVTagged recordsSetup guideNewParquet FolderColumnar filesSetup guideNewSQLite FolderEmbedded databasesSetup guideAccessDesktop databaseSetup guideBigQueryCloud warehouseSetup guideDuckDBEmbedded analyticsSetup guideMariaDBRelational engineSetup guideMySQLRelational engineSetup guideOracleRelational engineSetup guidePostgreSQLRelational engineSetup guideSnowflakeCloud warehouseSetup guideSQL ServerRelational engineSetup guideSQLiteEmbedded fileSetup guideSupabaseHosted PostgresSetup guideAivenManaged · PG + MySQLSetup guideAlloyDBPostgres-compatibleSetup guideAmazon RDSManaged · multi-engineSetup guideAzure MySQLManaged MySQLSetup guideAzure PostgreSQLManaged PostgresSetup guideAzure SQLManaged SQL ServerSetup guideCloud SQLManaged · multi-engineSetup guideCrunchy BridgeManaged PostgresSetup guideDigitalOceanManaged · PG + MySQLSetup guideFly.ioManaged PostgresSetup guideHerokuManaged PostgresSetup guideIBM CloudManaged PostgresSetup guideKoyebServerless PostgresSetup guideMySQL HeatWaveManaged MySQLSetup guideNeonServerless PostgresSetup guideNileMulti-tenant PostgresSetup guideOCI PostgreSQLManaged PostgresSetup guideOVHcloudManaged · PG + MySQLSetup guidePlanetScaleManaged PostgresSetup guidePrismaServerless PostgresSetup guideRailwayManaged PostgresSetup guideRenderManaged PostgresSetup guideScalewayManaged · PG + MySQLSetup guideSkySQLManaged MariaDBSetup guideTimescaleTime-series PostgresSetup guideVultrManaged · PG + MySQLSetup guideXataServerless PostgresSetup guideGA4Web analyticsSetup guideGoogle AdsAdvertisingSetup guideHubSpotCRM & marketingSetup guideiTickMarket dataSetup guideSearch ConsoleSEO & searchSetup guideShipStationShipping & fulfillmentSetup guideShopifyE-commerceSetup guideStripePayments & billingSetup guide

No connector matches that search. Try a shorter term, or clear the filter.

What a connector does

One agent. Read-only SQL. Your credentials stay put.

A connector is a saved connection the Network Agent owns inside your network — not a tunnel we open into it.

Outbound only

The agent opens one encrypted connection out. Requests and results both travel over it — no inbound port, no VPN, no firewall change.

SELECT only, enforced

A validator on the agent accepts only SELECT, WITH and EXPLAIN. Credentials never leave your network — we never see them.

Results, anywhere

The same connector feeds Excel, Google Sheets, the Query Builder, scheduled reports, a REST endpoint and Nova — without duplicating the connection.

Hosted & managed

Paste the string your provider gave you.

Managed services are branded cards in the wizard, so you pick the provider you actually pay — then it connects over the standard wire protocol underneath.

01

Pick your provider

Supabase, Neon, RDS, Cloud SQL, Azure — 27 managed services have their own card with the right defaults, ports and pooler settings already known.

02

Copy the connection details

Take them from your provider’s dashboard. The card already knows the shape that provider hands you — which fields it maps to, and where the non-standard ones live.

03

Query it like any database

Underneath it is PostgreSQL, MySQL, MariaDB or SQL Server, so every feature — filters, scheduling, sharing, Nova — works identically.

One dialect

Join a database to a folder, or an API, in one query

A CSV drop-folder, a hosted Postgres, and Stripe sit in the same FROM clause. File Set tables carry a source-file column so you can query the whole set or one file.

revenue_by_account.sqlSQL
-- Supabase (Postgres) joined to Stripe, one read-only query
SELECT   a.name, SUM(s.total) AS mrr
FROM     public.accounts a
JOIN     stripe.subscriptions s ON s.customer = a.stripe_id
WHERE    s.status = 'active'
GROUP BY a.name
ORDER BY mrr DESC
Results stream into Excel and Google Sheets read-only, always
Get started

Connect your first data source in minutes.

Install the agent, connect a database or point it at a folder, run a query. If it is SQL — or a folder of files that should be — it is already supported.

Try the Free Tier Team Pricing

Free forever tier · No credit card required

How it works

    Connect this source Query Streams home