Log inSubmit server
Data Studio logo

Data Studio

Connects AI coding agents to SQL and NoSQL databases through local SqlKit and DocKit desktop bridges.

InstallRepositoryWebsite
Full README Compare alternatives Embed widget Claim this listing

Install Data Studio

Pick your client, then add it in one click or paste the config.

STDIONode.js
~/Library/Application Support/Claude/claude_desktop_config.json
claude_desktop_config.json
{
  "mcpServers": {
    "data-studio": {
      "command": "npx",
      "args": [
        "-y",
        "@geek-fun/data-studio-mcp"
      ]
    }
  }
}

Settings → Developer → Edit Config, merge this into mcpServers, then fully quit and reopen Claude Desktop.

Install not yet sandbox-tested— we haven't run this command through our automated sandbox yet. Not a red flag; we're working through the catalog.

Overview

Data Studio MCP server connects AI coding agents to databases through the local dockit and sqlkit desktop apps. SqlKit supports more than 70 SQL databases, while dockit provides access to Elasticsearch, OpenSearch, MongoDB, and DynamoDB. Agents can inspect schemas, query data, explain SQL, and perform permitted write or administrative operations through one stdio MCP configuration. Reach for Data Studio when database credentials should remain inside local desktop applications and access needs separate read, write, and destructive-operation controls.

Use cases

  • Explore schemas and relationships across connected SQL databases
  • Query Elasticsearch, MongoDB, DynamoDB, or OpenSearch in plain language
  • Explain SQL queries and inspect database performance details
  • Control database writes and destructive actions with permission modes

Key features

  • Routes SQL and NoSQL operations through one MCP server
  • Supports more than 70 SQL databases through SqlKit
  • Connects to Elasticsearch, OpenSearch, MongoDB, and DynamoDB
  • Keeps credentials inside local desktop applications
  • Provides read-only, read/write, and full-access permission modes
  • Runs locally over stdio and localhost bridges

Tools80

What Data Studio lets your agent do. Expand a tool for its parameters.

Self-reported~679 tokens of context

Security & health

What it can reach and how reliably it runs — from our own checks, not the maintainer’s claims.

ActiveLast checked 3d ago

Recent health check succeeded.

7/7 checks healthy over the last 45d

Supply-chain signal

No high-severity advisories surfaced by our automated scan.

Critical 0High 0Medium 0Low 0

Scanned 3d ago via OSV.dev · @geek-fun/data-studio-mcp (npm)

Adoption & maintenance

Factual signals from GitHub, npm, and our automated checks — not a rating.

GitHub stars3Stargazers on the source repository.
npm downloads251Package downloads in the last 30 days.
Last commit5d agoMost recent push to the default branch.
Tools exposed80Callable tools this server registers over MCP.

How Data Studio works

What Data Studio does

Data Studio gives MCP-compatible AI coding agents a single interface for working with SQL and NoSQL databases. It routes SQL operations through SqlKit and document-database operations through dockit, both running as local desktop applications.

The SQL side covers more than 70 database systems, including PostgreSQL, MySQL, SQL Server, Oracle, SQLite, DuckDB, ClickHouse, Snowflake, and BigQuery. The NoSQL side supports Elasticsearch, OpenSearch, MongoDB, and DynamoDB. A single MCP registration exposes both backends when both desktop applications are installed and configured.

How it works

The MCP package runs as a local stdio process. It communicates with the SqlKit and dockit bridges over localhost, while those applications manage database connections and credentials. Agents refer to connections by opaque connection IDs; database passwords and keys remain inside the desktop applications.

Operations are controlled by permission modes configured in the applications:

  • Read Only permits schema exploration and SELECT queries.
  • Data Read/Write adds operations such as INSERT, UPDATE, and index changes, while excluding deletes and drops.
  • Full Access permits all supported actions, including DELETE, DROP, and TRUNCATE.

Permissions can be adjusted per connection. Destructive actions require explicit confirmation, and SQL statements are classified before execution so tools reject operations outside their allowed statement category.

Setup and configuration

Install and launch dockit and/or sqlkit, create the required database connections, and enable Settings → MCP Bridge → Auto-start. Installing both applications provides the combined SQL and NoSQL tool set.

Install the MCP package globally:

Terminal
npm install -g @geek-fun/data-studio-mcp

Alternatively, run it with npx:

Terminal
npx -y @geek-fun/data-studio-mcp

Register a stdio MCP server with npx and the argument -y @geek-fun/data-studio-mcp. The repository includes setup examples for Codex, Claude Code, Cursor, Windsurf, and OpenCode. The applications support macOS, Windows, and Linux.

Tools and capabilities

The server exposes tools for listing connections and checking bridge status. SqlKit tools cover database, schema, table, view, procedure, function, trigger, index, foreign-key, session, and connection discovery. They also support table inspection, query explanation, object DDL retrieval, slow-query inspection, SQL execution, write operations, deletion, DDL, session termination, and privilege changes.

The dockit backend exposes Elasticsearch search, document retrieval, and index listing operations. Tool names use the data_studio__{backend}__{action} structure, making the selected backend explicit.

Limitations and notes

Database credentials are not passed through MCP, but the agent receives broad access to data permitted by the configured policies. The bridge binds only to 127.0.0.1, so it is intended for local use rather than hosting as a remote database gateway. Read-only mode is the default, and destructive operations may require a confirmation prompt from the MCP client.

Read the full README View source on GitHub

Reviews

No reviews yet — be the first to share how it worked for you.

Frequently Asked Questions about Data Studio

Data Studio is an MCP server that connects AI coding agents to SQL databases through SqlKit and to Elasticsearch, OpenSearch, MongoDB, and DynamoDB through dockit. Its main tools inspect schemas, query data, explain SQL, manage database objects, and apply permission-controlled writes or destructive operations.

Install it with `npm install -g @geek-fun/data-studio-mcp`, or run it directly with `npx -y @geek-fun/data-studio-mcp`. Launch dockit and/or sqlkit first, configure a database connection, and enable the MCP Bridge auto-start setting.

Yes, Data Studio works with Cursor through a stdio MCP entry. Add a `data-studio` server to `.cursor/mcp.json` for a project or `~/.cursor/mcp.json` globally, using `npx` with `-y @geek-fun/data-studio-mcp` as its arguments.

Data Studio keeps database credentials inside dockit and sqlkit. The MCP interface uses opaque connection IDs, while the local bridge binds to `127.0.0.1`; passwords and keys are not sent as MCP tool arguments.

Yes, Data Studio can prevent or confirm destructive operations through its permission modes. Read Only blocks writes, Data Read/Write excludes deletes and drops, and Full Access permits them; DELETE, DROP, and TRUNCATE actions can require explicit user confirmation.

Related MCP servers

All in Databases All alternatives
Add a badge to your READMEShow it’s listed — live SVG, Markdown or HTML.

AllMCPs Directory Badge

Full Badge Customizer

Showcase your server listing on GitHub or your project documentation. Embed this dynamic SVG badge to highlight official listing status and live engagement.

Badge Style:
Live Dynamic SVG PreviewData Studio AllMCPs Directory Badge
Markdown (GitHub README)
[![AllMCPs](https://allmcps.com/api/badge/data-studio?style=directory)](https://allmcps.com/mcp/data-studio)
HTML Embed
<a href="https://allmcps.com/mcp/data-studio"><img src="https://allmcps.com/api/badge/data-studio?style=directory" alt="Data Studio on AllMCPs" /></a>