proffesor-for-testing__agentic-qe/docs/examples/phase1

Phase 1 Examples (v1.0.5)

This directory contains code examples demonstrating Phase 1 features: Multi-Model Router and Streaming MCP Tools.

---

📁 Directory Structure

phase1/
├── routing/
│   ├── basic-routing.ts           # Basic model selection
│   └── cost-tracking.ts            # Cost tracking and reporting
├── streaming/
│   ├── test-generation-stream.ts   # Streaming test generation
│   └── test-execution-stream.ts    # Streaming test execution
└── cost-tracking/
    └── budget-management.ts        # Budget management and forecasting

---

🚀 Quick Start

Prerequisites

# Install dependencies
npm install agentic-qe@latest
npm install progress chalk ora

# Set API keys
export OPENAI_API_KEY="sk-..."
export ANTHROPIC_API_KEY="sk-ant-..."

Run Examples

# Routing examples
npx ts-node routing/basic-routing.ts
npx ts-node routing/cost-tracking.ts

# Streaming examples
npx ts-node streaming/test-generation-stream.ts
npx ts-node streaming/test-execution-stream.ts

# Budget management
npx ts-node cost-tracking/budget-management.ts

---

📚 Examples Overview

1. Basic Routing (routing/basic-routing.ts)

Demonstrates basic model selection based on task complexity.

Features:

• Simple task routing (GPT-3.5)
• Complex task routing (GPT-4)
• Critical task routing (Claude Sonnet 4.5)
• Cost estimation

Key Concepts:

const router = new ModelRouter(config);
const selection = await router.selectModel(task);
console.log(`Selected: ${selection.modelId}`);
console.log(`Cost: $${selection.estimatedCost}`);

---

2. Cost Tracking (routing/cost-tracking.ts)

Comprehensive cost tracking and reporting.

Features:

• Real-time cost tracking
• Budget management
• Cost breakdown by model/agent
• Savings analysis
• CSV/JSON export

Key Concepts:

const tracker = new CostTracker();
tracker.setBudget({ period: 'daily', limit: 50.00 });

// Track execution
tracker.record(execution);

// Get report
const report = await tracker.getBreakdown();
console.log(`Savings: ${report.savings.percentage}%`);

---

3. Test Generation Streaming (streaming/test-generation-stream.ts)

Real-time progress updates for test generation.

Features:

• Live progress bar
• Test-by-test updates
• Coverage tracking
• Colored terminal output
• Metrics monitoring

Key Concepts:

const stream = await fleet.streamTestGeneration(options);

stream.on('progress', (update) => {
  console.log(`${update.progress}%`);
});

stream.on('test:generated', (test) => {
  console.log(`✓ ${test.name}`);
});

const result = await stream.complete();

---

4. Test Execution Streaming (streaming/test-execution-stream.ts)

Real-time progress updates for test execution.

Features:

• Live test results
• Pass/fail tracking
• Coverage updates
• Suite summaries
• Failed test details

Key Concepts:

const stream = await fleet.streamTestExecution(options);

stream.on('test:passed', (test) => {
  console.log(`✓ ${test.name}`);
});

stream.on('test:failed', (test) => {
  console.error(`✗ ${test.name}: ${test.error.message}`);
});

---

5. Budget Management (cost-tracking/budget-management.ts)

Comprehensive budget management and forecasting.

Features:

• Multi-level budgets (daily, monthly, project, team)
• Alert configuration
• Dynamic adjustments
• Cost forecasting
• Report generation

Key Concepts:

const budgetManager = new BudgetManager();

// Set budgets
await budgetManager.setBudget({
  period: 'daily',
  limit: 50.00,
  onExceeded: 'pause'
});

// Check status
const status = await budgetManager.getStatus('daily-budget');

// Forecast
const forecast = await budgetManager.forecastCosts({
  period: 'month'
});

---

🎯 Use Cases

Cost Optimization

1. Start with Basic Routing to understand model selection
2. Add Cost Tracking to monitor spending
3. Implement Budgets to control costs
4. Use Forecasting for planning

Real-Time Monitoring

1. Enable Streaming for long operations
2. Add Progress Bars for better UX
3. Track Metrics for performance
4. Handle Errors gracefully

Team Management

1. Set Team Budgets for cost allocation
2. Configure Alerts for notifications
3. Generate Reports for stakeholders
4. Optimize Rules based on usage

---

📖 Related Documentation

• Multi-Model Router Guide
• Streaming API Tutorial
• Cost Optimization Best Practices
• Migration Guide
• Routing API Reference
• Streaming API Reference

---

💡 Tips

Performance

• Use streaming for operations >5 seconds
• Enable caching for repeated operations
• Batch operations when possible
• Monitor memory usage

Cost Savings

• Use GPT-3.5 for simple tasks (10x cheaper)
• Enable adaptive routing (learns over time)
• Set daily/monthly budgets
• Review cost reports weekly

Reliability

• Always handle errors
• Implement fallback chains
• Clean up event listeners
• Monitor connection status

---

🐛 Troubleshooting

"No models available"

// Ensure models are configured
const router = new ModelRouter({
  models: {
    available: [
      { id: 'gpt-3.5-turbo', provider: 'openai', costPer1kTokens: 0.002, maxTokens: 4096 }
    ]
  }
});

"API key not found"

export OPENAI_API_KEY="sk-..."
export ANTHROPIC_API_KEY="sk-ant-..."

"No progress updates"

// Attach listeners BEFORE starting stream
stream.on('progress', handler);
await stream.start();  // Not before!

---

🤝 Contributing

Have a great example? Submit a PR!

1. Create example file in appropriate directory
2. Add documentation comments
3. Include expected output
4. Update this README

---

Questions? Open an issue: https://github.com/proffesor-for-testing/agentic-qe/issues