AI Agent Automation Platform

A modular TypeScript platform for building, testing, and orchestrating specialized AI agents. The platform is designed to support scalable agent development, automation workflows, and future integrations with Microsoft AI technologies and cloud services.

Overview

The AI Agent Automation Platform provides the foundation for developing specialized AI agents that automate software engineering, testing, documentation, and development workflows.

This project follows a modular architecture that allows agents to be developed, tested, and deployed independently while sharing common utilities and orchestration logic.

Current Features

• Modular project structure
• TypeScript development environment
• GitHub Actions CI workflow
• Jenkins CI pipeline
• Environment configuration
• Documentation framework
• Unit testing structure
• Coordinator Agent
• Manual Test Generation Agent
• Email Agent
• Microsoft Agent Framework integration
• Microsoft Semantic Kernel integration
• Microsoft Graph service foundation
• Azure AI Foundry service foundation
• Express REST API
• Health check endpoint
• REST API routing
• Browser-based Web User Interface

Planned AI Agents

Implemented

• Coordinator Agent
• Manual Test Generation Agent
• Email Agent

Planned

• Code Review Agent
• Documentation Agent

Future Enhancements

The platform is designed to support future AI capabilities and integrations, including:

• Microsoft Copilot Studio
• GitHub Pull Request Automation
• Retrieval-Augmented Generation (RAG)
• Azure OpenAI Service

Microsoft AI Services

Current service foundations:

• Microsoft Agent Framework
• Microsoft Semantic Kernel
• Microsoft Graph
• Azure AI Foundry

Development Environment

Recommended Tools

• Node.js 20+
• VS Code
• Git
• ESLint
• Prettier

Install dependencies:

npm install

Technology Stack

• TypeScript
• Node.js
• Express
• Jenkins
• GitHub Actions
• Git
• REST APIs
• AI-assisted development tools

Jenkins CI Pipeline

This project includes a Jenkins Declarative Pipeline for continuous integration.

Pipeline Stages

1. Checkout – Clone the source code from GitHub.
2. Install Dependencies – Run npm ci.
3. Lint – Run npm run lint.
4. Build – Run npm run build.
5. Test – Run npm test.

Running the Pipeline

1. Open Jenkins.
2. Select the ai-agent-automation-platform pipeline.
3. Click Build Now.
4. Review the Console Output.

A successful build finishes with:

Finished: SUCCESS

A failed build finishes with:

Finished: FAILURE

Getting Started

Install dependencies:

npm install

Build the project:

npm run build

Run unit tests:

npm test

Run ESLint:

npm run lint

Create a local environment configuration:

cp .env.example .env

For complete setup instructions, see:

• Local Development Setup

---

REST API

The AI Agent Automation Platform exposes a REST API for executing AI agents.

Start the API

Development:

npm run dev

Production:

npm start

The API runs at:

http://localhost:3000

Endpoints

GET /

Serves the browser-based Web UI.

GET /health

Returns the API health status.

Example:

{
  "status": "UP"
}

GET /api/agents

Returns API information.

Example:

{
  "message": "Agent API"
}

POST /api/agents/execute

Routes a prompt to the appropriate AI agent using the Coordinator Agent.

Example request:

{
  "prompt": "Generate manual test cases for login"
}

Example response:

{
  "success": true,
  "result": {
    "selectedAgent": "manual-test",
    "message": "Request routed to the Manual Test Agent."
  }
}

REST API Flow

Client
   │
HTTP Request
   │
Express Server
   │
Middleware
   │
Routes
   │
Controller
   │
Coordinator Agent
   │
Routing Response

---

Web User Interface

The AI Agent Automation Platform includes a browser-based interface for interacting with AI agents.

Features

• Select an AI agent
• Enter a prompt
• Execute requests through the REST API
• View JSON responses
• Loading indicator
• Error handling

Running the Web UI

Start the application:

npm run dev

or

npm start

Open your browser:

http://localhost:3000

Example

Selected Agent

Coordinator Agent

Prompt

Generate manual test cases for login

Example Response

{
  "success": true,
  "result": {
    "selectedAgent": "manual-test",
    "message": "Request routed to the Manual Test Agent."
  }
}

Web UI Flow

Browser
    │
    ▼
Web UI
    │
    ▼
REST API
    │
    ▼
Coordinator Agent
    │
    ▼
Specialized AI Agent

---

Testing

The application can be verified using:

• Vitest
• Supertest
• curl
• Postman Desktop

Run all automated tests:

npm test

Run linting:

npm run lint

Build the project:

npm run build

---

Documentation

• Project Structure
• Architecture Documentation
• Setup Documentation

---

Project Status

🚧 This project is currently under active development.

Current capabilities include:

• Coordinator Agent
• Manual Test Generation Agent
• Email Agent
• REST API
• Browser-based Web UI
• GitHub Actions CI
• Jenkins CI Pipeline
• Microsoft Agent Framework integration
• Microsoft Semantic Kernel integration
• Microsoft Graph service foundation
• Azure AI Foundry service foundation

Future iterations will introduce additional AI agents, workflow orchestration, drag-and-drop automation, Microsoft AI integrations, and automated software engineering capabilities.