Spaces:
Sleeping
Sleeping
metadata
title: Gradio MCP Playground
emoji: 🛝
colorFrom: blue
colorTo: green
sdk: gradio
sdk_version: 4.31.0
app_file: app.py
pinned: true
license: apache-2.0
tags:
- mcp
- agents
- gradio
- hackathon
short_description: Build AI agents visually - No code required!
🛝 Gradio MCP Playground
🏆 Hugging Face MCP Hackathon Submission
This project democratizes AI agent development by providing a visual, no-code platform for creating, testing, and deploying MCP (Model Context Protocol) agents. Built with Gradio, it makes agent creation accessible to everyone - from beginners to experts.
✨ What Makes This Special?
- 🚀 Zero to Agent in 30 Seconds - Choose a template, customize, and deploy
- 🎨 Visual Builder - No coding required with our drag-and-drop interface
- 🤖 AI-Powered Assistance - Three specialized AI modes to guide you
- 🔌 Live MCP Integration - Connect and test real MCP servers instantly
- 📦 One-Click Deploy - Deploy to production with a single button
🎥 Demo Video
Watch our 2-minute demo to see the playground in action!
🌟 Key Features
🛝 AI Assistant Hub
Three specialized AI assistants help you at every step:
- Adam (General) - Your friendly AI companion with MCP superpowers
- Liam (MCP Agent) - Expert in MCP development and troubleshooting
- Arthur (Agent Builder) - Architect for complex agent systems
🏗️ Visual Server Builder
Create MCP servers without writing code:
- Template Gallery - Pre-built templates for common use cases
- Live Code Generation - See your Python code as you build
- Instant Testing - Test your tools without deployment
- Custom Templates - Start from scratch or modify existing ones
🔌 MCP Connections
Connect to any MCP server with ease:
- Quick Connect - One-click connection to popular servers
- Tool Discovery - Automatically discover available tools
- Live Testing - Execute tools and see results in real-time
- Activity Monitoring - Track all tool executions
📦 Server Management
Professional-grade server management:
- Dashboard View - Monitor all your servers at a glance
- One-Click Deploy - Deploy to Hugging Face Spaces instantly
- Version Control - Track changes and rollback if needed
- Share & Collaborate - Share your agents with the community
🚀 Quick Start
Try the Demo (No Installation)
Visit our Hugging Face Space to try the platform instantly!
Local Installation (Full Features)
# Clone repository
git clone https://github.com/seanpoyner/gradio-mcp-playground
cd gradio-mcp-playground
# Install dependencies
pip install -e ".[all]"
# Run the unified dashboard
gmp dashboard
# Or run with custom port
gmp dashboard --port 8081
💡 Use Cases
For Developers
- Rapid Prototyping - Test ideas without boilerplate code
- API Integration - Wrap any API as an MCP server
- Tool Development - Create reusable tools for AI assistants
For Businesses
- Custom Assistants - Build domain-specific AI agents
- Workflow Automation - Create multi-step processes
- Team Collaboration - Share tools across your organization
For Educators
- Teaching AI Concepts - Visual way to explain agent architectures
- Student Projects - Let students build without infrastructure worries
- Interactive Demos - Show real AI capabilities in action
For Researchers
- Protocol Testing - Experiment with MCP implementations
- Tool Benchmarking - Compare different approaches
- Rapid Experimentation - Test hypotheses quickly
📸 Screenshots
🛠️ Technical Implementation
Architecture
- Frontend: Gradio 4.31.0 with custom themes
- Backend: Python with async support
- Protocol: Model Context Protocol (MCP)
- Deployment: Hugging Face Spaces compatible
Key Technologies
- Gradio - Interactive web UI framework
- MCP - Standardized protocol for AI tools
- AsyncIO - High-performance async operations
- Pydantic - Data validation and serialization
🎯 Why This Matters
The Model Context Protocol (MCP) is powerful but can be complex to implement. Our playground removes these barriers:
- Accessibility - No need to understand protocol details
- Speed - Build in minutes instead of hours
- Learning - See how MCP works through examples
- Community - Share and discover agents easily
🤝 Contributing
We welcome contributions! See our Contributing Guide for details.
Development Setup
# Install in development mode
pip install -e ".[dev]"
# Run tests
pytest
# Format code
black .
ruff check .
📄 License
This project is licensed under the Apache License 2.0 - see the LICENSE file.
🙏 Acknowledgments
- Hugging Face - For hosting and the MCP Hackathon
- Gradio Team - For the amazing framework
- Anthropic - For the Model Context Protocol
- Community - For feedback and contributions
📞 Contact & Support
- GitHub Issues: Report bugs or request features
- Discussions: Join the conversation
- Twitter: @seanpoyner
Made with ❤️ for the Hugging Face MCP Hackathon 2025
Empowering everyone to build AI agents
Empowering everyone to build AI agents