Indexes the coding-agent sessions already sitting on your disk — months of history from before you installed it — and serves them back over MCP. 3.5 GB searched in ~1.5 ms, no LLM, no embeddings.
DBHub
Summary
A universal database MCP server — one integration for PostgreSQL, MySQL, SQL Server, SQLite, and MariaDB.
Features
- One server for Postgres, MySQL, SQL Server, SQLite, and MariaDB
- Token-efficient schema exposure for large databases
- Read-only mode and SSH tunnel support
Installation
Set up this MCP server in your favorite AI agent — copy a ready-made configuration below.
Description
DBHub is a single MCP server that speaks to every common relational database, which saves configuring and maintaining a different server per engine.
Supported databases
PostgreSQL, MySQL, MariaDB, SQL Server, and SQLite, all through one connection-string interface.
What makes it practical
- Token efficiency. Schema information is exposed compactly, which matters enormously on a database with hundreds of tables — naive schema dumps consume the entire context window before a query is written.
- Read-only mode, enforced at the server rather than trusted to the model.
- SSH tunnelling for databases that are not directly reachable, which is most production databases.
- Multiple transports — stdio for local clients, HTTP for remote deployment.
- A demo mode with a sample database, so you can evaluate it without pointing it at anything real.
Setup
npx @bytebase/dbhub --transport stdio --dsn "<connection string>". Built and maintained by Bytebase, whose main product is database change management, so the operational instincts show.
Related MCP Servers
Weaviate's MCP server, built into the database itself — hybrid search, schema inspection, tenant listing and batch upserts at /v1/mcp, scoped by RBAC.
Google's open-source MCP server for enterprise databases — one binary that fronts AlloyDB, BigQuery, Spanner, Postgres, MySQL, Oracle, MongoDB, Snowflake and more.
MongoDB's official MCP server — query databases, explore schemas, and manage Atlas clusters from an AI client.

