Narad MCP Server provides a multi-channel backend that integrates GitHub, Email, and WhatsApp to automate tasks via HTTP requests. The server exposes endpoints that accept incoming platform events or user commands and orchestrate AI-powered workflows across connected services. Real-time and offline operation modes enable flexible usage, while the secure Python/Flask backend handles routing, authentication, and task execution. Tools available include GitHub issue/PR interactions, email-based actions, and WhatsApp messaging automation, all orchestrated through Narad’s central controller (narad_server).

Prerequisites:

• Python 3.8+ (recommended 3.9+)
• pip (comes with Python)
• Optional: virtual environment tooling (venv)

Step-by-step:

1. Clone the repository: git clone <repository-url> cd <repo-directory>

2. Create and activate a virtual environment (recommended): python -m venv venv

   On Windows:

   venv\Scripts\activate

   On macOS/Linux:

   source venv/bin/activate

3. Install dependencies: pip install -r requirements.txt

4. Configure environment variables (examples): export NARAD_PORT=5000 export NARAD_GITHUB_TOKEN=your_github_token

5. Run the server: python -m narad_server

6. Optional: run with a specific host/port by setting environment variables or editing the code as needed.

Notes:

• Ensure network access to GitHub and any external services you plan to connect (Email/WhatsApp gateways).
• If a requirements.txt does not exist, install Flask and any dependencies listed in the project docs or setup files.

Tips and common considerations:

• Environment variables govern authentication and port configuration; keep tokens secure and do not expose them in logs.
• If you encounter Flask binding issues, verify the host and port settings and firewall rules.
• For WhatsApp integration, ensure you have the necessary gateway/WhatsApp Business API access configured as per the integration module requirements.
• Review theNarad modules (email_agent.py, github_agent.py, whatsapp_agent.py) for available endpoints and payload formats.
• Enable real-time mode for responsive interactions or offline mode for batch processing as needed.
• If the app uses a model loader, ensure the model_path or model sources are accessible at startup.