Connect AI to files and databases using a curated list of MCP servers
If you want AI to access files on your computer or external databases to perform tasks, you can find the necessary connections here. This repository is a list of Model Context Protocol (MCP) servers organized by topic, showing which tools and data AI can interact with. You can find the functionality you need and apply it immediately, or use it as a reference when developing new servers.
What MCP does
MCP is an open protocol that allows AI models to interact safely with local and remote resources. Through standardized servers, it enables AI to perform various tasks such as file access, database connections, and API integration. This list includes both servers ready for production environments and those in the experimental stage. Each server defines how AI accesses specific tools or data.
Criteria for choosing a server
Servers are finely categorized by use case and technology stack.
- Implementation language Servers are distinguished by the language they are written in, including Python, TypeScript, Go, Rust, C#, Java, C/C++, and Ruby.
- Service scope They are divided into cloud services, local services, and embedded systems to specify which environments AI interacts with.
- Operating system Icons indicate which platform the server runs on: macOS, Windows, or Linux.
Resources to help you get started
Beyond a simple list, it provides tutorials and community links needed to start with MCP. It includes practice-oriented guides such as the MCP quickstart, Claude Desktop app setup, and SQLite database integration. You can also share information with other developers through Reddit and Discord communities. A web-based directory is also provided, which stays in sync with the repository for easier navigation.
Things to know before using
This list targets cases where you install and run servers directly. If you are looking for hosted servers that connect only via URL, you should refer to other lists. You must verify the stability and security of each server in its individual repository, as this list covers both production-ready and experimental servers. While useful for developers wanting to connect specific tools or databases to AI, you should directly review the code quality and maintenance status of the servers themselves.
