Routing Director MCP Server

A comprehensive Model Context Protocol (MCP) Server and Client for Juniper's Routing Director with GPT-4 powered intelligent service management.

🚀 Overview

This project provides an intelligent interface to Juniper's Routing Director through MCP (Model Context Protocol), enabling natural language service creation, deletion, and management with GPT-4 powered analysis. The system includes both a server component and a user-friendly Streamlit web client.

✨ Key Features

🤖 GPT-4 Powered Intelligence

• Natural Language Processing: Create and delete services using simple English commands
• Intelligent Query Analysis: Automatic detection of service types and required operations
• Smart Service Routing: GPT-4 determines the best tool for each request

🛠️ Service Management

• L2 Circuit Services: Full lifecycle management with interactive forms
• L3VPN Services: Viewing and management (creation/deletion coming soon)
• EVPN Services: Viewing and management (creation/deletion coming soon)

🌐 Interactive Web Interface

• Streamlit Chat Interface: Natural conversation flow for service management
• Interactive Forms: User-friendly forms with default values for service configuration
• Real-time Status: Monitor service creation and deletion progress
• Configuration Preview: Review JSON configs and JUNOS CLI before deployment

🔧 Advanced Features

• 2-Step Workflow: Upload → Deploy for service creation
• 3-Step Deletion: Modify → Create Order → Execute for service deletion
• Bulk Operations: Handle multiple services efficiently
• Data Visualization: Enhanced tables with quality analysis and export options
• CLI Generation: Automatic JUNOS CLI configuration generation

📋 Prerequisites

• Python 3.8 or higher
• Access to Juniper Routing Director
• OpenAI API key (for GPT-4 functionality)

🛠️ Installation

1. Clone the repository

   git clone <repository-url>

2. Install dependencies

   pip install -r requirements.txt

3. Create .env file Create a .env file in the root directory with your credentials:

   USERNAME=your_routing_director_username
   PASSWORD=your_routing_director_password
   ORG_ID=your_organization_id
   OPENAI_API_KEY=your_openai_api_key

🚀 Quick Start

Starting the Web Client

streamlit run mcpClient.py

The web interface will open at http://localhost:8501

Basic Usage

1. Connect to MCP Server: Click "🔌 Connect" in the web interface
2. Natural Language Commands: Type commands like:
   • "Create 2 L2 circuits from PNH-ACX7024-A1 to TH-ACX7100-A6 for customer SINET"
   • "Delete l2circuit1-135006"
   • "Show me all L3VPN services"

📚 Usage Examples

Service Creation Examples

Create an L2 circuit for customer SINET
Create L2 circuit from PNH-ACX7024-A1 to TH-ACX7100-A6 for customer SINET with service name test-l2ckt
Create 3 L2 circuits from PNH-ACX7024-A1 to TH-ACX7100-A6 for customer SINET

Service Deletion Examples

Delete l2circuit1-135006
Remove the L2 circuit service named test-circuit
Terminate service l2circuit1-135006

Service Viewing Examples

Show me all L3VPN services
Display L2 circuit instances
What EVPN services are running?
Show me the order history

🔧 Configuration

Environment Variables

Variable	Description	Example
USERNAME	Routing Director username	[email protected]
PASSWORD	Routing Director password	xyz
ORG_ID	Organization ID in Routing Director	123456789
OPENAI_API_KEY	OpenAI API key for GPT-4	sk-proj-eaVqS7tRMA...

Server Configuration

The MCP server runs on Routing Director API at https://x.x.x.x:48800 by default. Modify the BASE_URL in mcpServer.py if needed.

🏗️ Architecture

Components

1. mcpServer.py: Main MCP server with GPT-4 integration
2. mcpClient.py: Streamlit web client
3. l3vpn_parser.py: L3VPN service parser
4. l2ckt_parser.py: L2 Circuit service parser
5. l2vpn_evpn_parser.py: EVPN service parser
6. services_generator.py: Automatic service configuration generator
7. junos_cli_generator.py: JUNOS CLI configuration generator

Workflow

1. Query Analysis: GPT-4 analyzes natural language input
2. Tool Selection: System determines appropriate action
3. Form Generation: Interactive forms for missing details
4. Configuration: JSON and CLI configurations generated
5. Confirmation: User reviews configurations
6. Execution: 2-step deployment workflow
7. Monitoring: Real-time status updates

🛡️ Security

• Credentials stored in .env file (not in version control)
• Basic authentication with Routing Director
• HTTPS communication with API endpoints
• Input validation and error handling

🐛 Troubleshooting

Common Issues

1. Connection Failed

   • Verify .env credentials
   • Check network connectivity to Routing Director
   • Ensure Routing Director is accessible

2. GPT-4 Not Available

   • Verify OPENAI_API_KEY in .env
   • Check OpenAI API quota and billing

3. Service Creation Failed

   • Verify customer exists in Routing Director
   • Check device names in inventory
   • Review service configuration parameters

Debug Mode

Enable debug mode in the web interface to see:

• Session state information
• API request/response details
• Service generator status

📄 API Endpoints

The server provides access to these Routing Director endpoints:

• GET /order/instances - Service instances
• GET /order/orders - Service orders
• GET /order/customers - Customer information
• POST /order - Create service
• POST /order/.../exec - Execute service

📝 License

This project is licensed under the MIT License - see the LICENSE file for details.

🙋‍♂️ Support

For issues and questions:

1. Check the troubleshooting section
2. Review logs in the web interface
3. Use debug mode for detailed information
4. Create an issue in the repository

🔄 Version History

• v1.0.0: Initial release with GPT-4 integration
• Support for L2 Circuit creation and deletion
• Interactive web interface
• Natural language processing

---

Note: Remember to keep your .env file secure and never commit it to version control. The .env file should contain your actual credentials for Routing Director and OpenAI API access.