OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
MCP & Tooling

Dbhub

Dbhub (DBHub) is a TypeScript MCP server that acts as a gateway between MCP-compatible clients and PostgreSQL, MySQL, SQL Server, Oracle, MariaDB and SQLite databases. Its default tools support SQL execution and database-object discovery, while optional tools and custom operations extend the interface. It can run locally over stdio or as an HTTP server, and includes a browser-based Workbench.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Dbhub — OpenViST project profile in MCP & Tooling.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use DBHub when an MCP-compatible client needs a server-side interface to one or more supported relational databases, or when a browser Workbench is useful for running database tools and reviewing request traces.

Who it is for

  • Developers who want an MCP-compatible coding assistant to inspect schemas, validate queries or debug database data.
  • Teams that need one server interface for multiple supported database types or connections.
  • People providing non-technical colleagues with curated, read-only database views through an AI client.

Documented uses

  • Explore schemas, tables, columns, indexes and procedures through an AI client.
  • Run SQL queries through an MCP tool, with transaction support and safety controls.
  • Use a browser Workbench to run database tools and view request traces without an MCP client.
  • Expose reusable parameterized SQL operations as custom tools.
  • Connect multiple databases through one DBHub process using TOML configuration.
Before you choose

Requirements & limits

DBHub is open-source software under the MIT license. The documented NPM setup requires Node.js 22.5 or later; Docker is another documented installation route.

  • The supported database list includes PostgreSQL, MySQL, SQL Server, Oracle, MariaDB and SQLite.
  • Choose stdio when an AI tool will spawn DBHub as a local subprocess, or HTTP when running a standalone server.
  • Two tools are enabled by default; query-plan and health-check tools are opt-in, and custom parameterized operations can be configured.
  • Read-only mode, row limiting and query timeouts are documented controls; the demo quickstart notes that data changes require execute_sql not to be configured read-only.
  • DBHub's README distinguishes its minimal gateway from Bytebase's enterprise-level offering, which it describes as having approval flow, access control, data masking and audit logging.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Start DBHub in demo mode over HTTP

Run the documented demo server so an MCP client or browser can reach DBHub's sample database and Workbench.

Before you start

  • Node.js 22.5 or later for the NPM method, or Docker
  • An AI tool for the MCP-client workflow

Steps

  1. For NPM, run the command below; alternatively, run the command below.

    Documented command
    npx @bytebase/dbhub@latest --transport http --port 8080 --demo
    Documented command
    docker run --rm -p 8080:8080 bytebase/dbhub --transport http --port 8080 --demo

  2. Open the Workbench at http://localhost:8080/; the MCP endpoint is http://localhost:8080/mcp.

  3. Configure an AI tool using the relevant client-integration instructions, then try schema, table or sample employee-data questions.

Expected outcome

A local HTTP DBHub demo using a sample SQLite employee database, with a Workbench and MCP endpoint.

Guide limitations

  • This recipe covers demo mode with sample data, not configuration of a production database.
  • The quickstart lists client integration instructions but does not provide a single universal setup procedure for every AI tool.
Documentation 1 ↗
Source-based practical guide

Connect an AI client to a PostgreSQL database

Start DBHub over HTTP with a PostgreSQL DSN as the database connection.

Before you start

  • Node.js 22.5 or later for the NPM method
  • A PostgreSQL connection string
  • An MCP-compatible client configured to connect to DBHub

Steps

  1. Run the command below, substituting the connection details for your database.

    Documented command
    npx @bytebase/dbhub@latest --transport http --port 8080 --dsn "postgres://user:password@localhost:5432/dbname?sslmode=disable"

  2. Configure the MCP client to connect to DBHub using the HTTP transport and endpoint described in the installation and client-integration documentation.

Expected outcome

DBHub starts as an HTTP MCP server configured with a PostgreSQL connection.

Guide limitations

  • The command is an installation example; it does not establish that a particular database connection or client setup has been tested.
  • Connection-string formats and client-specific setup are covered in separate documentation.
Documentation 1 ↗Documentation 2 ↗
Evidence and transparency

Sources & review

Official repository ↗Guide documentation ↗
Source record details
Source availability
Active
Sources checked
9 October 2026
Original website title
bytebase/dbhub · Official homepage title · retained in its source language
How we review sourcesSuggest a correction