CLI-MCP is an open source Model Context Protocol (MCP) server that enables AI assistants to execute command line interface (CLI) commands within an isolated Docker container. Developed by kamikaze1120 and hosted on GitHub, the tool acts as a universal gateway for AI assistants such as Claude Desktop, Claude Code, and Cursor. Its primary function is to bridge the gap between AI assistants and command line environments, allowing users to run commands without manually switching to a terminal or copying context between applications. The tool operates by providing a single MCP server that replaces the need for multiple separate MCP servers. It supports over 50 different CLI tools and offers five distinct tools within one Docker container. Configuration is managed through a YAML file, with the server looking for specific configuration files in a defined order. Technical prerequisites include Python 3.10 or higher, Docker (either Desktop or CLI), and pip. For remote deployment, the server supports an HTTP transport and requires a CLI_MCP_AUTH_TOKEN for authentication. Typical use cases include asking an AI assistant to show recent git commits, run npm tests in a current project, check AWS S3 buckets, or format Python files with tools like Black. The workflow involves adding the server to an AI assistant's configuration file, such as claude_desktop_config.json for Claude Desktop or the MCP Servers settings in Cursor. Once configured, users can issue natural language requests to their AI assistant, which then executes the corresponding CLI commands inside the sandboxed Docker container. This approach aims to provide safe execution of commands by containing them within an isolated environment, reducing the potential impact of errors or unintended actions.