π Tabby-MCP-Server
@thuanpham582002
About π Tabby-MCP-Server
MCP server for control Tabby terminal
Config
Add this server to your MCP-compatible client using the configuration below.
{
"mcpServers": {
"tabby-mcp": {
"type": "sse",
"url": "http://localhost:3001/sse"
}
}
}Tools
4List available Tabby terminal sessions
Execute a shell command in a terminal tab
Read terminal buffer content
Retrieve full/paginated command output
Overview
What is π Tabby-MCP-Server?
A Tabby Terminal plugin that exposes your active terminal sessions through the Model Context Protocol (MCP), allowing MCP-compatible AI clients to discover terminal tabs, execute commands, read terminal buffers, and retrieve long command output.
How to use π Tabby-MCP-Server?
Install via the Tabby Plugin Store (Settings β Plugins) or by cloning the repository and running a Docker build with make build-dist followed by the provided install script. After restarting Tabby, MCP clients connect to the SSE endpoint at http://localhost:3001/sse using configuration such as {"mcpServers":{"tabby-mcp":{"type":"sse","url":"http://localhost:3001/sse"}}}.
Key features of π Tabby-MCP-Server
- Terminal session discovery β list local and SSH-backed Tabby tabs
- Command execution in a selected terminal tab
- Robust output capture with exit-code parsing
- Terminal buffer access β read visible and history ranges
- Long output pagination via
outputId - Optional pair programming mode with confirmation dialogs
Use cases of π Tabby-MCP-Server
- AI client runs shell commands in a live Tabby terminal session
- Developer retrieves full command output for debugging or analysis
- Read terminal buffer history without leaving the AI chat interface
- Automate repetitive terminal workflows through an MCP-compatible assistant
FAQ from π Tabby-MCP-Server
What are the requirements?
Tabby Terminal must be installed and running. Docker is required for local builds. The host needs an MCP-compatible client with SSE support (e.g., Claude Code, Cursor, Windsurf, Codex).
How do I install the plugin?
Use the Tabby Plugin Store (Settings β Plugins β βTabby MCPβ) or clone the repository, run make build-dist, and execute bash scripts/copy_to_plugin_folder.sh. Restart Tabby after installation.
What MCP clients are supported?
Any client that supports SSE MCP servers, such as Claude Code, Cursor, Windsurf, and Codex / OpenAI Codex CLI.
How do I configure the server?
Configure via Tabby under Settings β Plugins β MCP. Defaults include port 3001, debug logging, and pair programming mode with confirmation dialogs and terminal auto-focus.
What is pair programming mode?
An optional safety layer that shows confirmation dialogs before command execution, allows rejecting commands with feedback, and tracks command history. It also supports automatic terminal focus.
Frequently asked questions
What are the requirements?
Tabby Terminal must be installed and running. Docker is required for local builds. The host needs an MCP-compatible client with SSE support (e.g., Claude Code, Cursor, Windsurf, Codex).
How do I install the plugin?
Use the Tabby Plugin Store (Settings β Plugins β βTabby MCPβ) or clone the repository, run `make build-dist`, and execute `bash scripts/copy_to_plugin_folder.sh`. Restart Tabby after installation.
What MCP clients are supported?
Any client that supports SSE MCP servers, such as Claude Code, Cursor, Windsurf, and Codex / OpenAI Codex CLI.
How do I configure the server?
Configure via Tabby under Settings β Plugins β MCP. Defaults include port 3001, debug logging, and pair programming mode with confirmation dialogs and terminal auto-focus.
What is pair programming mode?
An optional safety layer that shows confirmation dialogs before command execution, allows rejecting commands with feedback, and tracks command history. It also supports automatic terminal focus.
Basic information
More Other MCP servers
Production-ready MCP integrations for AI applications
Klavis-AIKlavis AI: MCP integration platforms that let AI agents use tools reliably at any scale
ghidraMCP
LaurieWiredMCP Server for Ghidra
Website
FunnyWolfAdversary simulation and Red teaming platform with AI
Blender
ahujasidOpen-source MCP to use Blender with any LLM
MaxKB
1Panel-devπ₯ MaxKB is an open-source platform for building enterprise-grade agents. εΌΊε€§ζη¨ηεΌζΊδΌδΈηΊ§ζΊθ½δ½εΉ³ε°γ
Comments