💬 MCP Assistant Playground

https://github.com/user-attachments/assets/c60a59c0-5ff5-487e-be47-bdcc8253fbc8

https://github.com/user-attachments/assets/85e7453f-950d-41a4-a1f5-1abd3a587067

A Streamlit-based chatbot interface powered by OpenAI GPT-4o that intelligently routes user input to custom MCP tools such as GPT chat, image generation, Supabase queries, and text-to-speech.

Built for rapid experimentation with AI-powered tool routing, inspired by Claude-style confirmation flows.

✨ Features

• Natural language tool selection using GPT-4o
• MCP tool execution via fastmcp
• Real-time OpenAI image generation (DALL·E 3)
• Text-to-speech audio synthesis (GPT-4o mini TTS)
• Supabase integration for member CRUD operations
• Streamlit UI with tool result rendering (image/audio)

🛠️ Prerequisites

• Python 3.10+
• OpenAI API key
• Supabase project (optional, for member tools)

🧪 Getting Started

1. Clone the repository

git clone https://github.com/niawjunior/mcp-assistant-playground.git
cd mcp-assistant-playground

2. Set up virtual environment

python -m venv .venv
source .venv/bin/activate  # or `.venv\Scripts\activate` on Windows

3. Install dependencies

pip install -r requirements.txt

4. Set up environment variables

Create a .env file in the root directory with the following variables:

OPENAI_API_KEY=your_openai_api_key
SUPABASE_URL=your_supabase_url
SUPABASE_KEY=your_supabase_key

5. Run the application

python launch.py

6. Access the chat interface

Open your web browser and navigate to http://localhost:8501.