> ## Documentation Index
> Fetch the complete documentation index at: https://dbhub.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Intro to DBHub

![DBHub](https://mintcdn.com/dbhub/wULOyZshGvBbIq0K/images/logo/full-light.svg?fit=max\&auto=format\&n=wULOyZshGvBbIq0K\&q=85\&s=2f01abc06c649641ee05aa4e3b65378a)

DBHub is a minimal MCP server: token-efficient, zero-dependency, and just two tools by default with opt-in extras. This lightweight gateway allows MCP-compatible clients to connect to and explore different databases.

```
 +------------------+    +--------------+    +------------------+
 |                  |    |              |    |                  |
 |  Claude Desktop  +--->+              +--->+    PostgreSQL    |
 |                  |    |              |    |                  |
 |  Claude Code     +--->+              +--->+    SQL Server    |
 |                  |    |              |    |                  |
 |  Codex           +--->+    DBHub     +--->+    Oracle        |
 |                  |    |              |    |                  |
 |  Cursor          +--->+              +--->+    SQLite        |
 |                  |    |              |    |                  |
 |  VS Code         +--->+              +--->+    MySQL         |
 |                  |    |              |    |                  |
 |  Other Clients   +--->+              +--->+    MariaDB       |
 |                  |    |              |    |                  |
 |                  |    |              +--->+    ClickHouse    |
 |                  |    |              |    |                  |
 +------------------+    +--------------+    +------------------+
     MCP Clients           MCP Server             Databases
```

## Token Efficiency

DBHub loads just 2 tools by default at **0.5k tokens** — 16-25x fewer than alternatives — keeping the context window open for your actual work.

| MCP Server | Default Config | Default Tools |
| - | - | - |
| **DBHub** | **0.5k** | 2 (`execute_sql`, `search_objects`) |
| MCP Toolbox | 7.7k | 29 |
| Supabase MCP | 11.6k | 36 |

## Use Cases

* **Local Development**: Schema exploration, query validation, and data debugging with Claude Code, VS Code, Cursor, etc.
* **Non-Technical Access**: Expose curated, read-only views to non-technical staff via Claude Desktop, VS Code, Cursor, etc.
* **Multi-Database Consolidation**: Replace separate MCP servers for each database with a single DBHub process
* **Production Troubleshooting**: Read-only diagnostics with guardrails against runaway queries

> [!NOTE]
>
> Read-only mode is a guardrail, not a permission boundary. Always connect with a least-privilege, read-only database user. See [Read-Only Mode](https://dbhub.ai/config/read-only-mode.md).

## Supported Databases

PostgreSQL, MySQL, SQL Server, MariaDB, Oracle, ClickHouse, and SQLite.

## Why DBHub?

DBHub brings powerful database capabilities to AI coding assistants:

* **Minimal**: Zero dependency, token efficient with a minimal set of MCP tools to maximize context window
* **Multi-Database**: PostgreSQL, MySQL, MariaDB, SQL Server, Oracle, ClickHouse, and SQLite through a single interface
* **Multi-Connection**: Connect to multiple databases simultaneously with TOML configuration
* **Guardrails**: Read-only mode, row limiting, and query timeout to prevent runaway operations
* **Secure Access**: SSH tunneling and SSL/TLS encryption

## Key Features

### MCP Tools

DBHub provides AI assistants with direct database access through these tools:

* **[execute\_sql](https://dbhub.ai/tools/execute-sql.md)**: Run queries with transaction support and safety controls
* **[search\_objects](https://dbhub.ai/tools/search-objects.md)**: Explore schemas, tables, columns, indexes, and procedures
* **[explain\_sql](https://dbhub.ai/tools/explain-sql.md)** (opt-in): Show a query's execution plan without running it
* **[health\_check](https://dbhub.ai/tools/health-check.md)** (opt-in): Report connection pool state and buffer cache hit ratio
* **[Custom Tools](https://dbhub.ai/tools/custom-tools.md)**: Define reusable, parameterized SQL operations

### Workbench

A [web-based interface](https://dbhub.ai/workbench/overview.md) for running database tools and viewing request traces without requiring an MCP client.

## Architecture

```mermaid theme={null}
flowchart TB
    %% Human User
    User["Human User"]

    %% Workbench
    WB["DBHub Workbench"]

    %% AI Agent with MCP Client
    subgraph Agent["AI Agent"]
        MC["MCP Client"]
    end

    %% Configuration
    Config["dbhub.toml"]

    %% DBHub Server
    subgraph DBHub["DBHub Server"]
        MCP["/mcp endpoint"]
        subgraph Tools["MCP Tools"]
            ES["execute_sql"]
            SO["search_objects"]
            EX["explain_sql (opt-in)"]
            HC["health_check (opt-in)"]
            CT["custom tools"]
        end
    end

    %% Databases
    Databases["PostgreSQL, MySQL, SQL Server, MariaDB, Oracle, ClickHouse, SQLite"]

    %% Connections
    User --> WB
    WB --> DBHub

    Agent --> MC
    MC --> MCP
    MCP --> Tools

    Config -.->|loads| DBHub
    Tools --> Databases
```

DBHub acts as a gateway between databases and AI agents. AI agents access databases through the `/mcp` endpoint using MCP tools, while the Workbench provides direct browser-based access for humans.

> [card] [Installation](https://dbhub.ai/installation)
>
> Install DBHub using Docker, NPM, the MCP Bundle, or the Claude Code and Codex plugins

> [card] [Quickstart](https://dbhub.ai/quickstart.md)
>
> Get DBHub running with your first database connection


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.