pega

amribhatt/pega

3.2

If you are the rightful owner of pega and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to henry@mcphub.com.

The Model Context Protocol (MCP) server is designed to interface with the Pega Platform, providing tools and resources for efficient case management.

Tools
  1. Verify Connectivity

    Test connection to Pega Platform

  2. Get Case Types

    List available case types

  3. Create Case

    Create new cases with specified case type

Pega MCP + ADK Agent

A complete setup for Pega case management using MCP (Model Context Protocol) server and ADK (Agent Development Kit) agent.

Overview

This repository contains two main components:

  • Pega MCP Server (pega-mcp/) - Connects to Pega Platform APIs
  • Pega ADK Agent (pega-adk/) - AI agent that uses MCP tools to manage Pega cases

Quick Start

1. Clone the Repository

git clone https://github.com/amribhatt/pega.git
cd pega

2. Configure Environment

# Copy environment templates
cd pega-mcp
copy env.template .env

cd ../pega-adk
copy env.template .env

Edit the .env files with your actual credentials:

  • pega-mcp/.env - Pega Platform credentials
  • pega-adk/.env - ADK agent configuration

3. Create Virtual Environments

# Create virtual environment for MCP server
cd pega-mcp
python -m venv .venv
.venv\Scripts\activate  # On Windows
# source .venv/bin/activate  # On Linux/Mac

# Install MCP server dependencies
pip install -r requirements.txt

# Create virtual environment for ADK agent
cd ../pega-adk
python -m venv .venv
.venv\Scripts\activate  # On Windows
# source .venv/bin/activate  # On Linux/Mac

# Install ADK agent dependencies
pip install -r requirements.txt

4. Start the System

Terminal 1 - Start MCP Server:

cd pega-mcp
.venv\Scripts\activate  # On Windows
# source .venv/bin/activate  # On Linux/Mac
python server.py

Terminal 2 - Start ADK Agent:

cd pega-adk
.venv\Scripts\activate  # On Windows
# source .venv/bin/activate  # On Linux/Mac
adk run pega_adk_agent

Or to start with web interface:

cd pega-adk
.venv\Scripts\activate  # On Windows
# source .venv/bin/activate  # On Linux/Mac
adk web

Configuration

Required Pega Credentials

You'll need these from your Pega Platform:

  • Pega Platform URL - Your Pega instance URL
  • OAuth Client ID - OAuth2 client identifier
  • OAuth Client Secret - OAuth2 client secret
  • Application Alias - Your Pega application alias

Environment Files

Copy the template files and edit with your credentials:

pega-mcp/.env - MCP Server Configuration:

PEGA_BASE_URL=https://your-pega-instance.com
PEGA_CLIENT_ID=your_client_id
PEGA_CLIENT_SECRET=your_client_secret
APP_ALIAS=your_app_alias

pega-adk/.env - ADK Agent Configuration:

MCP_SERVER_URL=http://localhost:8080/mcp/
AI_MODEL=gemini-2.0-flash
AGENT_NAME=pega_adk_agent

Usage

MCP Server Features

The MCP server provides these tools:

  • Verify Connectivity - Test connection to Pega Platform
  • Get Case Types - List available case types
  • Create Case - Create new cases with specified case type

ADK Agent Features

The ADK agent can:

  • Connect to Pega Platform via MCP tools
  • List available case types
  • Create new cases
  • Provide natural language responses about case management

Example Interactions

User: "List available case types"
Agent: "Found 3 case types:
  1. Home Loan Application (ID: H-28021)
  2. Credit Card Application (ID: C-15045)
  3. Personal Loan Request (ID: P-89012)"

User: "Create a home loan case"
Agent: "Case created successfully with ID: LOAN-2024-001"

Project Structure

pega/
ā”œā”€ā”€ README.md               # This file
ā”œā”€ā”€ pega-mcp/              # MCP Server
│   ā”œā”€ā”€ server.py          # MCP server implementation
│   ā”œā”€ā”€ tools.py           # Pega API tools
│   ā”œā”€ā”€ resources.py       # MCP resources
│   ā”œā”€ā”€ requirements.txt   # MCP dependencies
│   ā”œā”€ā”€ env.template      # Environment template
│   └── dx-apis/          # API documentation
└── pega-adk/             # ADK Agent
    ā”œā”€ā”€ pega_adk_agent/
    │   └── agent.py      # ADK agent implementation
    ā”œā”€ā”€ requirements.txt   # ADK dependencies
    └── env.template      # Environment template

Troubleshooting

Common Issues

  1. "Missing configuration" error

    • Copy env.template to .env in both directories
    • Edit .env files with your actual credentials
    • Ensure all required Pega credentials are provided
  2. Connection failed

    • Verify your Pega Platform URL is correct
    • Check OAuth credentials are valid
    • Ensure network connectivity to Pega instance
  3. MCP server won't start

    • Check port 8080 is available
    • Verify virtual environment is activated
    • Verify all dependencies are installed
    • Check .env file exists in pega-mcp/ directory
  4. ADK agent can't connect

    • Ensure MCP server is running on localhost:8080
    • Check MCP server URL in pega-adk/.env
    • Verify virtual environment is activated
    • Verify ADK dependencies are installed
    • Use adk run pega_adk_agent or adk web to start the agent

Security Notes

  • .env files are excluded from Git (see .gitignore)
  • Never commit real credentials to version control
  • Keep your OAuth credentials secure
  • The setup script masks secret input for security

Development

Adding New Tools

To add new Pega API tools:

  1. Add function to pega-mcp/tools.py
  2. Register tool in pega-mcp/server.py
  3. Update agent instructions in pega-adk/pega_adk_agent/agent.py

Customizing the Agent

Edit pega-adk/pega_adk_agent/agent.py to:

  • Change AI model
  • Modify agent instructions
  • Add custom tools
  • Update agent behavior

ADK Commands

  • adk run pega_adk_agent - Start agent in terminal mode
  • adk web - Start agent with web interface
  • adk --help - Show all available commands

License

This project is for educational and development purposes. Ensure compliance with your organization's policies when using with production Pega systems.