mTarsier — Open-Source Tool to Manage MCP Servers Across All AI Clients
MCP360 has released mTarsier, a free and open-source desktop application that unifies Model Context Protocol (MCP) server management across all major AI clients from a single interface. Released on March 16, 2026, it addresses a growing pain point: as developers use multiple AI clients (Claude Desktop, Claude Code, Cursor, VS Code, Windsurf, ChatGPT Desktop, Gemini CLI, and more), managing MCP server configurations across scattered JSON files becomes error-prone and tedious.
mTarsier provides JSON validation with inline error feedback, a built-in marketplace to install MCP servers directly into any supported client, automatic backups before every configuration change, and team snapshots that export setups as .tsr files for collaboration. A CLI tool called tsr is also included for terminal-first workflows.
The tool runs on macOS, Windows, and Linux with no internet connection or account required, and currently supports 12+ AI clients. It's fully local-first — no telemetry, no cloud dependency.
As the MCP ecosystem expands rapidly, mTarsier fills a critical gap in the toolchain: managing the configuration layer that connects agents to their tools.
GitHub: https://github.com/mcp360/mTarsier
← Back to all articles
mTarsier provides JSON validation with inline error feedback, a built-in marketplace to install MCP servers directly into any supported client, automatic backups before every configuration change, and team snapshots that export setups as .tsr files for collaboration. A CLI tool called tsr is also included for terminal-first workflows.
The tool runs on macOS, Windows, and Linux with no internet connection or account required, and currently supports 12+ AI clients. It's fully local-first — no telemetry, no cloud dependency.
As the MCP ecosystem expands rapidly, mTarsier fills a critical gap in the toolchain: managing the configuration layer that connects agents to their tools.
GitHub: https://github.com/mcp360/mTarsier