Dataki for Claude

Connect Claude to your database

Add Dataki as a custom connector in Claude, sign in once, and Claude can query the sources you have connected: it writes the SQL, Dataki runs it read-only on its own side, and the answer comes back as a chart in the conversation. Your database password never reaches Claude.

  1. Claude
  2. Dataki MCP server
  3. Your database, read-only
Signs in with
Sign in with OAuth. No key to copy.
Works in
Claude on the web, the desktop app and the phone app.
You need
A Claude account and a Dataki account with at least one data source.
Server URL
https://dataki.ai/api/mcp
Connect Claude Go to the steps

01Set it up

4 steps, a couple of minutes

Dataki's connect page walks through the same steps with buttons, and creates a key wherever one is needed.

  1. 01

    Open Claude's connector settings

    Go to Settings → Connectors → Add custom connector in Claude.

  2. 02

    Name it Dataki and paste the server URL

    Remote MCP server URL 
    https://dataki.ai/api/mcp
  3. 03

    Click Add, then Connect, and sign in to Dataki

    Approve what Claude may reach. Claude gets its own credential, bound to your Dataki account, which you can revoke. If you have no data connected yet, Dataki hands you a link to connect a source first; it takes about a minute.

  4. 04

    Ask it something, in a new chat

    Start with What data can you see in Dataki?. A conversation keeps the tools it started with, so a chat that was already open will not see Dataki.

Only use the Claude desktop app? Install Dataki as an extension instead

Dataki's connect page can download a desktop extension: one file with a key for you already in it, so there is no sign-in. Open the file and Claude shows Dataki; click Install. It works in that app on that computer only, which is why the connector above comes first.

If Claude answers without Dataki after installing, open Settings → Extensions in Claude and make sure Dataki is switched on.

02Ask it

What to ask in any Claude chat

The assistant finds your sources, reads what the columns mean, writes the SQL and hands it to Dataki, which runs it read-only and returns the rows. The query is part of every answer.

How did revenue move each month this year, and which plans carried it?

Dataki runs the query against your warehouse and draws the chart from the rows that came back, not from numbers the model typed.

Which customers upgraded this quarter, and what did each one add?

Answered with a table as well as a sentence, so the figures can be checked line by line.

Save these two charts to a dashboard called Weekly review and share it with the team.

The dashboard lives in Dataki on a permanent link. It refreshes on its own; nobody has to ask again next week.

Tell me when daily orders fall below 100.

watch_metric runs the query on a schedule and emails an alert when the number crosses the line you set. list_schedules shows what is already being watched.

03What it can reach

Any source you connect to Dataki

Claude sees the sources your Dataki account can open, and one question can read more than one of them.

04Worth knowing

What Claude gets, and what it never does

Charts arrive in the conversation
Claude draws MCP Apps, so Dataki answers with a live view beside the chat: charts, tables and headline numbers drawn by the same components as a Dataki dashboard. Each chart is checked against its rows before you see it, and Add to dashboard keeps it.
It acts as you, and only as you
The connector holds no credential of its own. Anything you cannot reach in Dataki, Claude cannot reach either, and team restrictions, spend ceilings and the query log all still apply.
Under the hood
OAuth 2.1 with PKCE and dynamic client registration: Claude discovers where to send you, you sign in and approve the scopes, and it receives a revocable credential. Nothing is typed into a config file.

More on the server itself, and every tool it offers, on the MCP server page. How credentials, keys and the query log work: security and governance.

Dataki and Claude

How do I connect Claude to a PostgreSQL database?
Add a remote MCP server that can reach the database as a custom connector. With Dataki, you connect PostgreSQL to Dataki once with a read-only user, then add https://dataki.ai/api/mcp as a custom connector in Claude and sign in. Claude can then list your sources, read the schema and run read-only queries, and the answer arrives as a chart in the chat.
Does the Dataki connector work in the Claude mobile app?
Yes. A connector added to your Claude account follows you to Claude on the web, the desktop app and the phone app. The desktop extension, the alternative route, works in the desktop app on one computer only.
Does Claude get my database password?
No. Your database credentials are stored encrypted in Dataki and are never sent to Claude or to any model. Claude calls Dataki's MCP server, the server runs the SQL on its side with a read-only role, and only the rows that come back reach the conversation.
Can Claude change or delete data through Dataki?
No. Dataki's query tool is read-only: a statement that changes data is refused before it runs, and the database connection itself uses a role that cannot write. Claude can create and edit Dataki dashboards, which are Dataki's own records, never rows in your database.
Which databases can Claude query through Dataki?
Any source connected to Dataki: Google BigQuery, PostgreSQL, MySQL, Amazon Redshift, Supabase, Google Sheets and uploaded CSV or Excel files, including hosted databases such as Neon, Amazon RDS, Cloud SQL and PlanetScale. Google Analytics 4, Firebase, Search Console, Firestore and Stripe work through their exports to BigQuery. One question can read more than one source.

Using another assistant?

Connect Claude. Ask it something.

Connect a source to Dataki, add the server to Claude, and the first answer is one question away. Anything worth keeping becomes a dashboard with a link that stays live.

Connect Claude Stuck? Ask us