AgentForge AI

AgentForge AI is a production-ready multi-agent workflow platform designed for research, coding, project planning, and autonomous task execution. It provides a comprehensive framework for building scalable, intelligent applications with specialized agents that collaborate to solve complex problems.

Features

• Planner Agent: Orchestrates task decomposition and workflow planning
• Research Agent: Gathers and synthesizes information from multiple sources
• Coder Agent: Develops, reviews, and optimizes code implementations
• Reviewer Agent: Evaluates outputs and ensures quality standards
• Refiner Agent: Iteratively improves results based on feedback
• Report Agent: Generates comprehensive documentation and reports

Architecture

The AgentForge AI platform follows a modular, extensible architecture:

Planner → Research/Coder → Reviewer → Refiner → Report

Core Components

• Agents: Specialized task handlers for different workflow stages
• Memory Layer: Persistent context management and conversation history
• Tool Layer: Integration with external services and APIs
• Workflow Tracking: Execution monitoring and performance metrics
• Evaluation Engine: Quality assessment and metric tracking

Core Capabilities

• Memory Layer: Sophisticated context management with agent-specific memory stores
• Tool Integration: Extensible tool framework for web search, document analysis, and GitHub integration
• Workflow Tracking: Complete execution traces and performance monitoring
• Evaluation Engine: Metrics collection and quality assessment
• Docker Deployment: Container-ready for production environments
• CI/CD Pipeline: Automated testing and deployment workflows
• Test Coverage: Comprehensive test suite for reliability

Installation

Prerequisites

• Python 3.8 or higher
• pip package manager
• Virtual environment (recommended)

Quick Start

1. Clone the repository:

git clone https://github.com/itsananytripathi/AgentForge-AI.git
cd AgentForge-AI

2. Create a virtual environment:

python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install dependencies:

pip install -r requirements.txt

4. Configure environment variables:

cp env.example .env
# Edit .env with your API keys and configuration

5. Run the application:

python app.py

Docker Installation

docker-compose up -d

Usage

Basic Example

from agents import PlannerAgent, ResearchAgent, CoderAgent

planner = PlannerAgent()
researcher = ResearchAgent()
coder = CoderAgent()

# Define your task
task = "Research Python best practices and create example code"

# Execute workflow
plan = planner.execute(task)
research = researcher.execute(plan)
code = coder.execute(research)

Example Workflows

Research Workflow

Comprehensive information gathering and synthesis using multiple sources:

1. Receive research query
2. Search multiple sources (web, documents, GitHub)
3. Synthesize findings
4. Generate research report

See research_ml_fundamentals.md for detailed example.

Coding Workflow

End-to-end code development with review and refinement:

1. Receive specifications
2. Generate code
3. Review for quality
4. Refine based on feedback
5. Generate documentation

See coding_rest_api.md for detailed example.

Project Planning Workflow

Strategic planning and task decomposition:

1. Define project scope
2. Break down into phases
3. Plan resource allocation
4. Generate project timeline
5. Create execution roadmap

See planning_ecommerce_platform.md for detailed example.

Memory Layer

The memory layer provides sophisticated context management:

• Conversation Memory: Maintains chat history and context
• Agent Context Store: Specialized memory for each agent
• Task Memory: Tracks task-specific information and state
• Memory Manager: Orchestrates memory operations and cleanup

Tool Layer

Extensible tool integration framework:

• Web Search Tool: Real-time information retrieval
• Document Search Tool: Local and remote document analysis
• GitHub Analysis Tool: Repository and code analysis
• Tool Manager: Centralized tool orchestration

Workflow Tracking

Complete execution monitoring:

• Execution Traces: Detailed step-by-step execution logs
• Workflow Tracker: Overall workflow state management
• Performance Metrics: Execution time and resource tracking

Testing

Run the test suite:

pytest tests/ -v

Test coverage includes:

• Agent functionality
• Memory operations
• Tool integration
• Workflow tracking
• End-to-end integration tests

Logging

Comprehensive logging for debugging and monitoring:

• Configured in utils/logger.py
• Log files stored in logs/ directory
• Adjustable log levels (DEBUG, INFO, WARNING, ERROR)

Contributing

Contributions are welcome! Please follow these guidelines:

1. Fork the repository
2. Create a feature branch
3. Make your changes with clear commit messages
4. Add tests for new functionality
5. Submit a pull request

For detailed guidelines, see CONTRIBUTING.md.

Roadmap

Phase 1 (Current)

• Multi-agent orchestration
• Core workflow implementation
• Memory and tool layers

Phase 2

• Multi-LLM Support (OpenAI, Anthropic, Local models)
• Vector Database Integration
• Enhanced agent collaboration

Phase 3

• Autonomous Agent Collaboration
• Cloud Deployment Options
• Advanced Evaluation Metrics
• Custom Agent Framework

Future Enhancements

• Web UI Dashboard
• REST API Interface
• Real-time Monitoring
• Advanced Analytics

Acknowledgements

This project builds upon best practices in:

• Multi-agent systems
• Large language models
• Software architecture
• Workflow automation

Author

Anany Tripathi

• GitHub: https://github.com/itsananytripathi
• Email: anany.tripathivns@gmail.com

License

See LICENSE file for license information.