MCP Server

Your data,
wherever you work

Connect Sequel to Claude, Cursor, Windsurf, or any MCP-compatible AI client — and query your databases, explore schemas, and build analyses without leaving the tools your team already uses.

Works with every MCP-compatible client

Quick Start

Connect in 60 seconds

Pick your AI client and paste the config. No terminal, no extra installs — just your Sequel API key.

~/Library/Application Support/Claude/claude_desktop_config.json

{
  "mcpServers": {
    "sequel": {
      "url": "https://api.sequel.sh/mcp",
      "headers": {
        "Authorization": "Bearer sql_your_api_key"
      }
    }
  }
}
Replace sql_your_api_key with a key from Settings → API Keys. Then restart your AI client.
Security

Your permissions, everywhere

Sequel's MCP server enforces the same workspace-level access controls as the Sequel app. There's no way to query data you haven't been given access to.

API key auth

Create a key in Settings → API Keys. Keys are prefixed sql_ and can be revoked any time.

Org-scoped access

Each key is tied to one organization. The AI can only see connections and data your org has explicitly connected to Sequel.

Stateless & auditable

Every tool call is independent and logged. No persistent sessions, no state stored on the MCP server between calls.

FAQ

Frequently asked questions

What is an MCP server?

MCP (Model Context Protocol) is an open standard that lets AI clients — Claude, Cursor, VS Code Copilot, and others — call external tools during a conversation. Sequel's MCP server exposes your connected databases as tools, so any MCP-compatible AI can query your data directly.

Do I need to install anything?

No. Sequel's MCP server runs at api.sequel.sh/mcp — there's nothing to install or self-host. Just add a one-block config to your AI client pointing at that URL with your API key.

Which AI clients are supported?

Any client that supports MCP over HTTP works: Claude Desktop, Claude.ai, Claude Code, Cursor, Windsurf, VS Code (GitHub Copilot), Zed, Gemini, Codex CLI, and more. If it supports the MCP Streamable HTTP transport, it works with Sequel.

How do I get an API key?

Sign in to your Sequel account and go to Settings → API Keys. Create a new key, copy it, and paste it into your MCP config as the Bearer token. Keys are prefixed sql_ and can be revoked any time.

Which data sources can the AI access?

The AI can access any data source you've connected in Sequel — PostgreSQL, MySQL, ClickHouse, Google Analytics, Google Search Console, and more. It can only see connections your organization has explicitly added.

Is my data sent to the AI provider?

Query results are returned to the AI as context within your conversation, subject to that provider's data handling policies. Sequel never stores query results permanently — they exist only for the duration of the tool call.

Can multiple team members use the same MCP server?

Yes. Each team member creates their own API key from Settings → API Keys and adds it to their personal AI client config. The MCP server is stateless — it handles any number of concurrent users.

Start querying from any AI tool

Generate a free API key and connect Sequel to your AI client in under a minute.