agentops-ai

AgentOps: AI-Powered Requirements-Driven Test Automation

AgentOps is a next-generation, AI-powered development platform that automatically generates requirements and tests from your codebase. Using advanced multi-agent AI systems, AgentOps bridges the gap between technical implementation and business requirements through bidirectional analysis.

🚀 Quick Start

Installation

# Install from PyPI
pip install agentops-ai

# Or install from source
git clone https://github.com/knaig/agentops_ai.git
cd agentops_ai
pip install -e .

Basic Usage

# 1. Initialize your project
agentops init

# 2. Run complete analysis on your code
agentops runner myfile.py

# 3. Check results and status
agentops status

# 4. View traceability matrix
agentops traceability

That's it! AgentOps will automatically:

• ✅ Analyze your code structure
• ✅ Generate business-focused requirements
• ✅ Create comprehensive test cases
• ✅ Execute tests and provide results
• ✅ Generate traceability matrices

📋 Available Commands

🚀 Power User Options

Combined Usage:

# Batch process entire project with automation
agentops runner --all --auto-approve -v

# Generate tests for all files
agentops tests --all --auto-approve

# Detailed analysis with verbose output
agentops analyze myfile.py -v

🎯 Enhanced CLI Experience

Interactive Onboarding

New users can get started quickly with the interactive onboarding:

agentops onboarding

This will:

• ✅ Check your installation and API key setup
• ✅ Create sample files for demonstration
• ✅ Run a complete demo workflow
• ✅ Show you all available commands and options
• ✅ Guide you through next steps

Batch Processing

Process your entire project with a single command:

# Analyze all Python files in your project
agentops runner --all

# Generate tests for all files
agentops tests --all

# With automation and verbose output
agentops runner --all --auto-approve -v

CI/CD Integration

Perfect for automated workflows:

# Fully automated analysis
agentops runner --all --auto-approve

# Check project status
agentops status -v

🔧 Configuration

Environment Setup

Create a .env file in your project root:

# Required: OpenAI API Key
OPENAI_API_KEY=your_openai_api_key_here

# Optional: Custom settings
AGENTOPS_OUTPUT_DIR=.agentops
AGENTOPS_LOG_LEVEL=INFO

API Key Validation

Validate your API keys before running:

# Run the validation script
./scripts/validate_api_keys.sh

📁 Generated Artifacts

After running AgentOps, you'll find these artifacts in your project:

.agentops/
├── requirements/
│   ├── requirements.gherkin    # Requirements in Gherkin format
│   └── requirements.md         # Requirements in Markdown format
├── tests/
│   ├── test_main.py           # Generated test files
│   └── test_coverage.xml      # Test coverage reports
├── traceability/
│   └── traceability_matrix.md # Requirements-to-tests mapping
└── reports/
    └── analysis_report.json   # Detailed analysis results

🎯 Key Features

🤖 Multi-Agent AI System

• Code Analyzer: Deep code structure and dependency analysis
• Requirements Engineer: Business-focused requirement extraction
• Test Architect: Comprehensive test strategy design
• Test Generator: High-quality, maintainable test code
• Quality Assurance: Automated test validation and scoring

📊 Requirements-Driven Testing

• Bidirectional Analysis: Code → Requirements → Tests
• Business Context: Requirements focus on real-world value
• Traceability: Complete mapping from requirements to tests
• Multiple Formats: Gherkin (BDD) and Markdown outputs

🔄 Streamlined Workflow

• Single Command: agentops runner does everything
• Automatic Export: Requirements and traceability always up-to-date
• Error Recovery: Robust handling of common issues
• Progress Tracking: Real-time status updates

📖 Examples

Basic Python File Analysis

# myfile.py
def calculate_total(items, tax_rate=0.1):
    """Calculate total with tax for a list of items."""
    subtotal = sum(items)
    tax = subtotal * tax_rate
    return subtotal + tax

def apply_discount(total, discount_percent):
    """Apply percentage discount to total."""
    discount = total * (discount_percent / 100)
    return total - discount

# Run analysis
agentops runner myfile.py

Generated Requirements (Gherkin):

Feature: Shopping Cart Calculations

Scenario: Calculate total with tax
  Given a list of items with prices [10, 20, 30]
  And a tax rate of 10%
  When I calculate the total
  Then the result should be 66.0

Scenario: Apply discount to total
  Given a total amount of 100
  And a discount of 15%
  When I apply the discount
  Then the final amount should be 85.0

Generated Tests:

def test_calculate_total_with_tax():
    items = [10, 20, 30]
    result = calculate_total(items, tax_rate=0.1)
    assert result == 66.0

def test_apply_discount():
    result = apply_discount(100, 15)
    assert result == 85.0

🛠️ Advanced Usage

Custom Configuration

# agentops_ai/agentops_core/config.py
LLM_MODEL = "gpt-4"
LLM_TEMPERATURE = 0.1
OUTPUT_DIR = ".agentops"
LOG_LEVEL = "INFO"

Python API

from agentops_ai.agentops_core.orchestrator import AgentOrchestrator

# Create orchestrator
orchestrator = AgentOrchestrator()

# Run analysis
result = orchestrator.run_workflow("myfile.py")

# Access results
print(f"Requirements: {len(result.requirements)}")
print(f"Tests generated: {len(result.test_code)} lines")
print(f"Quality score: {result.quality_score}")

🔍 Troubleshooting

Common Issues

API Key Errors:

# Set your OpenAI API key
export OPENAI_API_KEY="your-key-here"

Permission Errors:

# Ensure write permissions
chmod 755 .agentops/

Import Errors:

# Reinstall dependencies
pip install -r requirements.txt

Getting Help

# Show detailed help
agentops help

# Check version
agentops version

# Validate setup
agentops status

📚 Documentation

• Quick Start Guide
• Architecture Overview
• API Reference
• Integration Guide
• Changelog

🤝 Contributing

We welcome contributions! Please see our Contributing Guide for details.

📄 License

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

🆘 Support

• Issues: GitHub Issues
• Documentation: Project Docs
• Email: agentops_info@protonmail.com

---

AgentOps: Bridging technical implementation and business requirements through AI-powered analysis. 🚀