🎬 Demo

✨ Why AgentX?

AgentX is a GPU-accelerated, cross-platform desktop application that brings AI agents to your workflow. Built with cutting-edge technologies, it provides a seamless experience for interacting with multiple AI agents, editing code, managing tasks, and more—all in one unified interface.

🎯 Features

• 🤖 Multi-Agent Support - Connect and chat with multiple AI agents simultaneously via Agent Client Protocol (ACP)
• 💬 Real-time Conversations - Streaming responses with support for thinking blocks and tool calls
• 📝 Built-in Code Editor - LSP-enabled editor with syntax highlighting and autocomplete
• 🖥️ Integrated Terminal - Execute commands without leaving the app
• 🎨 Customizable Dock System - Drag-and-drop panels to create your perfect workspace
• 🌍 Internationalization - Support for multiple languages (English, 简体中文)
• 🎭 Theme Support - Light and dark themes with customizable colors
• 📊 Session Management - Organize conversations across multiple sessions
• 🔧 Tool Call Viewer - Inspect agent tool executions in detail
• 💾 Auto-save - Never lose your work with automatic session persistence
• ⚡ GPU-Accelerated - Blazing fast UI powered by GPUI framework

🤖 Supported Agents

Based on config.json in this repository we test the following agents:

Codex	Claude	Kimi Code
Qwen	Qoder	OpenCode
Gemini	AugmentCode	Iflow

More ACP-Compatible Agents

From the ACP "Agents implementing the Agent Client Protocol" list:

• AgentPool
• Blackbox AI
• Code Assistant
• Docker's cagent
• fast-agent
• GitHub Copilot (public preview)
• Goose
• JetBrains Junie (coming soon)
• Minion Code
• Mistral Vibe
• OpenHands
• Pi (via pi-acp adapter)
• Stakpak
• VT Code

📦 Installation

📥 Download Latest Release

# Extract and run
# Or double-click setup.exe to install

# Using winget (coming soon)
# winget install AgentX

# For Debian/Ubuntu (.deb)
sudo dpkg -i agentx_0.5.0_amd64.deb

# For other distros (.tar.gz)
tar -xzf agentx-v0.5.0-x86_64-linux.tar.gz
cd agentx
./agentx

# Or using AppImage
chmod +x agentx-v0.5.0-x86_64.AppImage
./agentx-v0.5.0-x86_64.AppImage

# Double-click .dmg and drag AgentX to Applications folder

# Using Homebrew (coming soon)
# brew install --cask agentx

🚀 Quick Start

1. Download AgentX for your platform from the releases page
2. Install following your OS-specific instructions above
3. Launch AgentX
4. Configure your AI agent in Settings → MCP Config
5. Start chatting with your agent!

🛠️ Development

# Clone the repository
git clone https://github.com/sxhxliang/agent-studio.git
cd agent-studio

# Build and run
cargo run

# Release build
cargo build --release

# Run with logging
RUST_LOG=info cargo run

# Run tests
cargo test

# Check code
cargo clippy

# Format code
cargo fmt

🏗️ Built With

• GPUI - GPU-accelerated UI framework from Zed Industries
• gpui-component - Rich UI component library
• Agent Client Protocol - Standard protocol for agent communication
• Tokio - Async runtime
• Tree-sitter - Syntax highlighting
• Rust - Memory-safe systems programming language

📖 Documentation

• User Guide - Learn how to use AgentX
• Architecture - Technical architecture and design
• Contributing - How to contribute to the project
• Agent Configuration - Set up your AI agents

❓ QA

Q: What should I do if no agents appear in the agent list?

A: This is usually caused by network access restrictions. Configure your proxy first, then set it on the startup page or in the Settings panel, and try again.

Q: Does the app manage agent authorization and availability?

A: No. AgentX provides the desktop studio experience only, and does not manage provider authorization state or service availability. Ensure your target agent is authorized and reachable before using the app.

Q: AgentX starts but no agent can respond. What should I check?

A: Open Settings -> MCP Config and verify your provider settings in config.json (API endpoint, key, command path, environment variables).

Q: How do I reset a broken dock layout?

A: Close AgentX and delete docks-agentx.json, then relaunch the app.

Q: Where can I find runtime/session data?

A: Layout/session runtime files are written under agentx/ , and session data is stored in sessions/.

Q: How can I debug startup or integration issues?

A: Run AgentX with logs enabled:

RUST_LOG=info cargo run

🤝 Contributing

We welcome contributions! Whether it's bug reports, feature requests, or pull requests—every contribution helps make AgentX better.

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

🌟 Show Your Support

If you find AgentX helpful, please consider:

• ⭐ Star this repository to show your support
• 🐦 Share it with your friends and colleagues
• 🐛 Report bugs to help us improve
• 💡 Suggest features you'd like to see

📝 License

This project is licensed under the Apache-2.0 License. See LICENSE file for details.

🙏 Acknowledgments

Special thanks to:

• Zed Industries for the amazing GPUI framework
• GPUI Component contributors
• All our contributors and supporters

Star History