Agentic Prompt Orchestration Architecture

A flexible, extensible framework for orchestrating multiple AI agents to solve complex tasks through coordinated prompt chaining and intelligent workflow management.

🚀 Features

• Multi-Agent Coordination: Specialized agents (Planner, Executor, Validator, Router, Synthesizer)
• 4 Execution Strategies: Sequential, Parallel, Conditional, and DAG (Directed Acyclic Graph)
• Dynamic Task Generation: Agents can create follow-up tasks based on outputs
• Template-Based Prompts: Flexible prompt templates with variable substitution
• Agent Memory System: Short-term, long-term, and working memory for each agent
• Validation Framework: Custom validation rules for quality assurance
• Dependency Resolution: Automatic task ordering based on dependencies
• Fully Typed: Complete TypeScript implementation with comprehensive types

📋 Prerequisites

• Node.js 18.x or 20.x
• npm 9.x or higher
• TypeScript 5.x

🛠️ Installation

# Clone the repository
git clone https://github.com/NeuralBlitz/agentic-prompt-orchestration.git
cd agentic-prompt-orchestration

# Install dependencies
npm install

# Build the project
npm run build

🏃 Quick Start

import { createOrchestrationSystem } from './index';

// Initialize the system
const { templateEngine, orchestrator, agents } = createOrchestrationSystem();

// Load workflow configurations
const workflowConfig = require('./workflows.json');
templateEngine.loadFromJSON(workflowConfig);

// Create and register agents
const planner = agents.createPlanner({
  id: 'planner-agent',
  role: 'planner',
  name: 'Task Planner',
  description: 'Plans and organizes tasks',
  capabilities: ['planning']
});

orchestrator.registerAgent(planner);

// Execute a workflow
const context = {
  conversationId: 'session-001',
  variables: { topic: 'AI Research' },
  history: [],
  state: {}
};

const results = await orchestrator.executeWorkflow('research-workflow', context);
console.log(results);

📖 Documentation

• Complete Guide - Comprehensive documentation with examples
• Usage Guide - Practical patterns and best practices
• Architecture - System design and component interactions

🧪 Testing

# Run all tests
npm test

# Run tests with coverage
npm test -- --coverage

# Run tests in watch mode
npm test -- --watch

🔧 Development

# Run in development mode
npm run dev

# Run linter
npm run lint

# Format code
npm run format

# Build for production
npm run build

📁 Project Structure

├── types.ts                    # Type definitions
├── PromptTemplateEngine.ts     # Template rendering engine
├── agents.ts                   # Agent implementations
├── orchestrator.ts             # Workflow orchestration
├── index.ts                    # Main entry point
├── example.ts                  # Example application
├── workflows.json              # Workflow configurations
├── advanced-workflow.json      # Advanced workflow example
├── __tests__/                  # Test files
│   └── PromptTemplateEngine.test.ts
├── .github/
│   └── workflows/
│       └── ci.yml              # GitHub Actions CI/CD
└── README.md                   # This file

🔄 CI/CD Pipeline

This project uses GitHub Actions for continuous integration and deployment:

• Build: Compiles TypeScript and creates distribution files
• Test: Runs unit tests with coverage reporting
• Lint: Checks code quality with ESLint and Prettier
• Multi-Node: Tests against Node.js 18.x and 20.x

🤝 Contributing

1. Fork the repository
2. Create a feature branch (git checkout -b feature/amazing-feature)
3. Commit your changes (git commit -m 'Add amazing feature')
4. Push to the branch (git push origin feature/amazing-feature)
5. Open a Pull Request

📝 Example Workflows

Research Pipeline

const researchWorkflow = {
  id: 'research-pipeline',
  strategy: { type: 'dag', config: {} },
  tasks: [
    { id: 'plan', agentId: 'planner' },
    { id: 'research', agentId: 'researcher', dependencies: ['plan'] },
    { id: 'validate', agentId: 'validator', dependencies: ['research'] },
    { id: 'synthesize', agentId: 'synthesizer', dependencies: ['validate'] }
  ]
};

Content Creation

const contentWorkflow = {
  id: 'content-creation',
  strategy: { type: 'parallel', config: { maxConcurrency: 3 } },
  tasks: [
    { id: 'intro', agentId: 'writer-1' },
    { id: 'body', agentId: 'writer-2' },
    { id: 'conclusion', agentId: 'writer-3' },
    { id: 'edit', agentId: 'editor', dependencies: ['intro', 'body', 'conclusion'] }
  ]
};

🐛 Troubleshooting

Common Issues

Issue: “Dependencies lock file is not found”

• Solution: Run npm install to generate package-lock.json

Issue: “Agent not found”

• Solution: Ensure agents are registered before executing workflows

Issue: “Circular dependency detected”

• Solution: Review task dependencies to ensure they form a DAG

📜 License

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

🙏 Acknowledgments

• Built with TypeScript for type safety
• Inspired by multi-agent AI systems research
• Designed for extensibility and real-world use cases

📧 Contact

• GitHub: @NeuralBlitz
• Issues: GitHub Issues

---

Built with ❤️ for the AI agent community