anyquery
Query anything (GitHub, Notion, +40 more) with SQL and let LLMs (ChatGPT, Claude) connect to using MCP
claude mcp add --transport stdio julien040-anyquery anyquery mcp --stdio
How to use
Anyquery provides an MCP (Model Context Protocol) server that lets LLMs and other clients interact with Anyquery's data querying capabilities. The MCP server is started from the Anyquery CLI and supports standard stdio or socket-based connections via the host/port option. To initialize the MCP server in the simplest way, run the MCP mode via the CLI and connect over stdio, which is suitable for embedded LLM clients. If you prefer connecting over a network, you can start the MCP server with an HTTP/SSE tunnel by using the host and port options and then point your client at that endpoint. Once running, LLMs can request context, execute SQL-style queries against files, databases, apps, or integrations that Anyquery supports, and can leverage built-in or plugin-backed data sources.
The Anyquery MCP server exposes capabilities for running SQL-like queries across a variety of data sources, managing plugins/integrations, and optionally bridging to LLMs for enhanced data access. You can connect LLMs via the MCP protocol, or use client tooling that supports function calling or HTTP-based MCP endpoints. The project emphasizes plug-in integrations to extend data sources (files, databases, apps, etc.) and can act as a MySQL-compatible server for certain clients when configured accordingly. To use it with an LLM, start the MCP server and provide the appropriate connection details to the LLM client or integration guide described in Anyquery's documentation.
How to install
Prerequisites:
- A supported operating system (Linux, macOS, or Windows) with internet access
- Basic command-line tools (bash, zsh, or PowerShell) and a package manager if you choose an install method
Option 1: Install via Homebrew (macOS/Linux with Homebrew)
brew install anyquery
Option 2: Install via APT (Debian/Ubuntu)
echo "deb [trusted=yes] https://apt.julienc.me/ /" | sudo tee /etc/apt/sources.list.d/anyquery.list
sudo apt update
sudo apt install anyquery
Option 3: Install via YUM/DNF (RHEL/CentOS/Fedora)
echo "[anyquery]
name=Anyquery
baseurl=https://yum.julienc.me/
enabled=1
gpgcheck=0" | sudo tee /etc/yum.repos.d/anyquery.repo
sudo dnf install anyquery
Option 4: Windows (Winget or Chocolatey)
winget install JulienCagniart.anyquery
# or
choco install anyquery
Option 5: Binary download
- Visit the Anyquery releases page: https://github.com/julien040/anyquery/releases
- Download the appropriate binary for your platform and add it to your PATH
After installation, you can verify by running:
anyquery --version
Then start the MCP server (see mcp_config to run in MCP mode).
Additional notes
Tips and notes:
- The MCP server is intended for LLM clients and tooling that support the MCP protocol. If your client expects HTTP/SSE, use the --host/--port options when launching Anyquery in MCP mode and point the client to that endpoint.
- Plugins: Anyquery uses a plugin-based model to extend data sources. You can install integrations from the official registry to expose new data sources to the MCP server.
- When running the MCP server, ensure network accessibility if you plan to connect from remote LLMs or services.
- If you enable the MySQL server feature within Anyquery, ensure you configure the appropriate port and access controls to avoid unauthorized access.
- Consult Anyquery documentation for details on supported data sources, plugin installation, and security considerations.
Related MCP Servers
netdata
The fastest path to AI-powered full stack observability, even for lean teams.
mindsdb
Query Engine for AI Analytics: Build self-reasoning agents across all your live data
excelize
Go language library for reading and writing Microsoft Excel™ (XLAM / XLSM / XLSX / XLTM / XLTX) spreadsheets
pgmcp
An MCP server to query any Postgres database in natural language.
facebook-ads-library
MCP Server for Facebook ADs Library - Get instant answers from FB's ad library
hyperterse
The MCP framework. Connect your data to your agents.