Why VibeShell

Most SSH clients were built in an era before AI. They have clunky interfaces, zero automation, and treat every session as a purely human activity.

VibeShell is different. It's a modern terminal with first-class support for AI coding agents — Claude Code, OpenAI Codex, Cursor, Windsurf, and any MCP-compatible tool. Through a built-in MCP server and auto-installable skill system, your AI agents can SSH into servers, execute commands, transfer files, and manage infrastructure — all while you watch in a beautiful UI or let it run autonomously.

Think of it as the terminal client that both you and your AI actually enjoy using.

Features

🤖 AI Agent Integration

The killer feature. VibeShell ships with a built-in MCP (Model Context Protocol) server and a skill installer that teaches AI agents how to use it.

What your AI can do	MCP Tool
List & manage servers	server_list, server_add, server_get
Open SSH sessions	session_create, session_attach
Execute remote commands	exec
Browse & transfer files	sftp_ls, sftp_upload, sftp_download
Create directories, move/delete files	sftp_mkdir, sftp_mv, sftp_rm

One-click install — VibeShell auto-detects Claude Code, Codex CLI, Cursor, Windsurf, and other MCP-compatible tools on your machine, then installs its skill (SKILL.md + MCP config) so your agent knows exactly how and when to use SSH.

You: "Deploy the latest build to production"
Agent: [Uses VibeShell] → connects to prod server → pulls code → restarts service → confirms deployment

🖥️ Terminal

• Multi-tab sessions — Connect to multiple servers simultaneously
• WebGL rendering — Buttery-smooth terminal via Xterm.js
• Local shell — Built-in local terminal alongside SSH sessions
• Ghost text completions — Fish-shell-style inline suggestions
• Keyboard shortcuts — Ctrl+N new server, Ctrl+K quick command, Ctrl+W close tab

🔐 SSH & Security

• SSH tunneling — Local forward, remote forward, and SOCKS5 dynamic forwarding
• Jump host (ProxyJump) — Connect through bastion/jump servers
• SSH agent forwarding — Forward local SSH agent to remote hosts
• Host key verification — TOFU with SHA-256 fingerprint management
• Encrypted credential storage — AES-encrypted passwords and key passphrases

📁 File Management

• SFTP browser — Built-in file manager with upload, download, preview, edit
• Drag & drop — Upload files by dragging into the SFTP panel
• Compression — Compress/extract archives on remote servers
• File preview — View text files, images, and code with syntax highlighting

⚡ Productivity

• Command snippets — Save, tag, and reuse frequently used commands
• Session recording — Record terminal sessions for playback and audit
• Post-login commands — Auto-execute commands after SSH connection
• Quick command palette — Ctrl+K to search and run anything
• Server groups & tags — Organize hundreds of servers effortlessly

🎨 Design

• Tokyo Night theme — A beautiful dark theme, not a generic UI
• Internationalization — English and 简体中文
• Native performance — Rust backend, ~8 MB binary, instant startup
• Cross-platform — Windows, macOS, Linux with native look and feel

Installation

Download

Platform	Download
Windows x64	.exe installer · .msi
macOS Apple Silicon	.dmg
macOS Intel	.dmg
Linux x64	.deb · .AppImage · .rpm

Build from Source

# Prerequisites: Node.js 18+, Rust 1.70+, Tauri prerequisites
git clone https://github.com/veithly/vibeshell.git
cd vibeshell
npm install
npx tauri dev      # Development mode with hot reload
npx tauri build    # Build release binary

Architecture

┌──────────────────────────────────────────────────────┐
│                     Frontend                          │
│     React 18 + Zustand + Xterm.js + Tailwind CSS     │
├──────────────────────────────────────────────────────┤
│                  Tauri IPC Bridge                     │
├──────────────────────────────────────────────────────┤
│                      Backend                          │
│         Rust + russh + rusqlite + tokio               │
│  ┌──────┐ ┌──────┐ ┌────────┐ ┌─────────┐ ┌─────┐  │
│  │ SSH  │ │ SFTP │ │ Tunnel │ │ Storage │ │ MCP │  │
│  └──────┘ └──────┘ └────────┘ └─────────┘ └─────┘  │
│                                              ▲       │
│  ┌──────────────────────────────────────────┐│       │
│  │        Skill Installer & Detector        ││       │
│  │  Claude Code · Codex · Cursor · ...      │┘       │
│  └──────────────────────────────────────────┘        │
└──────────────────────────────────────────────────────┘

• Frontend — React handles UI, Zustand manages state, Xterm.js renders terminals
• Backend — Rust provides native SSH/SFTP via russh, SQLite for persistence
• MCP Server — Exposes 16 tools for AI agents to manage servers, sessions, and files
• Skill Installer — Auto-detects AI tools and configures the VibeShell MCP skill

Development

npx tauri dev                                      # Dev server with hot reload
npx tsc --noEmit                                   # Type-check frontend
cargo check --manifest-path src-tauri/Cargo.toml   # Check Rust backend
cargo test --manifest-path src-tauri/Cargo.toml    # Run Rust tests

Project Structure

vibeshell/
├── src/                    # React frontend
│   ├── components/         # UI components (Terminal, SFTP, Settings, ...)
│   ├── stores/             # Zustand state stores
│   ├── i18n/               # Internationalization (en, zh)
│   └── lib/                # Utilities (tauri.ts, utils.ts)
├── src-tauri/              # Rust backend
│   └── src/
│       ├── commands/       # Tauri command handlers (60+ commands)
│       ├── ssh/            # SSH client & fingerprints (russh)
│       ├── sftp/           # SFTP operations
│       ├── tunnel/         # SSH tunneling engine
│       ├── mcp/            # MCP server & 16 tool definitions
│       ├── install/        # AI tool detector & skill installer
│       ├── storage/        # SQLite database & encrypted credentials
│       └── logging/        # Session recording
├── cli/                    # CLI companion tool
└── .github/workflows/      # CI + auto-release pipeline

Contributing

Contributions are welcome!

1. Fork the repository
2. Create your feature branch (git checkout -b feature/amazing-feature)
3. Commit your changes (git commit -m 'Add amazing feature')
4. Push to the branch (git push origin feature/amazing-feature)
5. Open a Pull Request

License

MIT — Made with care by veithly.