Mcpverse

🚀 MCPverse Project Documentation

MCPverse is a semantic search engine designed to discover Model Configuration Protocol (MCP) servers on GitHub. It indexes public MCP repositories and leverages vector embeddings and OpenSearch to enable powerful natural - language search.

📡 Online Demo

Try the hosted version of MCPverse: 👉 https://mcpverse.streamlit.app

✨ Features

• 🔍 Conduct semantic searches using OpenAI's text - embedding - ada - 002 model.
• 📚 Index names, descriptions, and README contents.
• ⚙️ Provide client configuration previews (if available).
• 🎨 Offer an interactive UI implemented with Streamlit.

📦 Project Structure

.
├── Backend
│   ├── fetch_repos.py      # Fetch MCP server repositories from GitHub.
│   ├── extract_config.py   # Logic to extract MCP client configurations from READMEs.
│   ├── embedder.py         # Logic for OpenAI and OpenSearch indexing and searching.
│   ├── github_scraper.py   # Search for MCP servers using the GitHub API.
│   └── data/
│       └── mcpverse_data.json  
├── Frontend
│   └── app.py               # Streamlit application.
├── requirements.txt
├── README.md
├── .env
├── .gitignore

📦 Installation

1. Clone the Repository

git clone https://github.com/Harika-BV/MCPverse.git
cd MCPverse

2. Set Up the Python Environment

python -m venv venv
source venv/bin/activate  # For Windows users: venv\Scripts\activate
pip install -r requirements.txt

3. Configure Environment Variables

Create a .env file in the root directory:

GH_API_KEY=your-github-token
ENV=local
OPENAI_API_KEY=your-openai-key
OPENSEARCH_HOST=localhost
OPENSEARCH_PORT=9200
OPENSEARCH_USER=admin
OPENSEARCH_PASS=admin

💡 You can also connect to your own hosted OpenSearch or Elasticsearch cluster.

4. Run OpenSearch Locally (Optional)

If you want to run OpenSearch locally:

docker run -d --name opensearch -p 9200:9200 \
  -e "discovery.type=single-node" \
  -e "plugins.security.disabled=true" \
  opensearchproject/opensearch:2.11.1

5. Index MCP Repositories

cd Backend
python embedder.py

This command will:

• Load MCP server data from GitHub (from mcpverse_data.json).
• Generate OpenAI embeddings.
• Index them into OpenSearch.

6. Run the Streamlit Application

cd Frontend
streamlit run app.py

Then open http://localhost:8501 in your browser.

🧑‍💻 Maintainer

Developed with ❤️ by Harika B V

⚠️ Disclaimer

All repositories and data are publicly available on GitHub.
MCPverse is a community project and has no association with any third - party MCP maintainers.