Unified Gmail account creation using both LLM intelligence and RPA automation
This project implements two parallel approaches for automated Gmail account creation:
- π€ Kilocode LLM: Intelligent form filling using Large Language Models (OpenAI/Claude)
- π§ RPA BitBrowser: Traditional browser automation with antidetect profiles
- LLM-Powered: AI-driven form filling with intelligent responses to dynamic challenges
- RPA-Based: Traditional browser automation with BitBrowser antidetect profiles
- Hybrid Controller: Intelligent routing between approaches based on success rates and requirements
- Intelligent Routing: Automatically selects optimal approach based on success rates and complexity
- Failover Support: Falls back to alternative approach if primary method fails
- Real-time Dashboard: Web-based monitoring and control interface
- Comprehensive Testing: Full integration test suite for both systems
- Performance Monitoring: Detailed statistics and cost tracking
- Node.js 14+
- BitBrowser application (for RPA automation)
- OpenAI or Claude API key (for LLM automation)
# Clone repository
git clone <repository-url>
cd "gmail kilocode creator bybitbot"
# Install dependencies
npm install
# Configure API keys (optional)
npm run kilocode:config -- --openai-key "your-api-key"npm run kilocode:status# Basic LLM automation
npm run kilocode:llm
# With options
npm run kilocode:llm -- --country DE --age "young adult" --provider openai# Basic RPA automation
npm run kilocode:rpa
# Multiple profiles
npm run kilocode:rpa -- --profiles 3 --country US# Intelligent routing
npm run kilocode:hybrid
# Force specific strategy
npm run kilocode:hybrid -- --strategy llm --count 5 --failovernpm run kilocode:dashboard
# Access at http://localhost:3000βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ
β Unified Dashboard β
β Web Interface + Real-time WS β
βββββββββββββββββββββββ¬ββββββββββββββββββββββββββββββββββββββββ
β
βββββββββββββββββββββββΌββββββββββββββββββββββββββββββββββββββββ
β Hybrid Controller β
β Intelligent Routing + Fallback β
βββββββββββββββ¬ββββββββββββββββββββββββββββββββ¬ββββββββββββββββ
β β
βββββββββββΌββββββββββ βββββββββββΌββββββββββ
β Kilocode LLM β β RPA Engine β
β - OpenAI/Claude β β - BitBrowser β
β - Smart Forms β β - Playwright β
β - Dynamic AI β β - Antidetect β
βββββββββββββββββββββ βββββββββββββββββββββ
| Requirement | LLM Approach | RPA Approach |
|---|---|---|
| Speed | β‘ Fast (API calls) | π Slower (browser) |
| Cost | π° API costs | π° Low operational |
| Complexity | π§ Handles dynamic | π§ Structured only |
| Reliability | π Depends on API | π― Browser-dependent |
| Scalability | β¬οΈ High | β¬οΈ Limited by profiles |
# LLM Configuration
OPENAI_API_KEY=your_openai_key
CLAUDE_API_KEY=your_claude_key
# BitBrowser Configuration
BITBROWSER_PORT=35100
BITBROWSER_API=http://127.0.0.1:35100
# Dashboard Configuration
DASHBOARD_PORT=3000
DASHBOARD_WS_PORT=3001{
"openaiApiKey": "your-api-key",
"claudeApiKey": "your-claude-key",
"bitbrowserPort": 35100,
"defaultStrategy": "intelligent",
"failoverEnabled": true,
"rateLimiting": {
"requestsPerMinute": 60,
"tokensPerMinute": 90000
}
}# Unified CLI
npm run kilocode # Show help
npm run kilocode:status # Check system health
npm run kilocode:config # Configure settings
# Automation Commands
npm run kilocode:llm # LLM automation
npm run kilocode:rpa # RPA automation
npm run kilocode:hybrid # Hybrid automation
npm run kilocode:dashboard # Start dashboard
# Testing Commands
npm run kilocode:test # Run all tests
npm run test:llm # Test LLM only
npm run test:rpa # Test RPA only
npm run test:hybrid # Test hybrid only# BitBrowser Management
npm run create-profiles # Create BitBrowser profiles
npm run list-profiles # List existing profiles
npm run check-bitbrowser # Check BitBrowser status
# Automa Workflows
npm run automa:simple # Simple Automa workflow
npm run automa:full # Full Automa workflow# Let the system choose the best approach
npm run kilocode:hybrid -- \
--strategy intelligent \
--country "DE" \
--count 3 \
--failover# Use LLM for complex requirements
npm run kilocode:llm -- \
--country "France" \
--age "middle aged" \
--gender "female" \
--provider claude \
--debug# Use RPA for bulk creation
npm run kilocode:rpa -- \
--profiles 10 \
--country "US" \
--headless# Start dashboard and create accounts via web interface
npm run kilocode:dashboard
# Navigate to http://localhost:3000- Real-time Statistics: Success rates, costs, execution times
- Strategy Comparison: Performance metrics for LLM vs RPA
- Live Account Creation: Create accounts via web interface
- Health Monitoring: System status and component health
- Activity Logs: Detailed execution logs and error tracking
# Example output from npm run kilocode:status
π Checking System Status...
π€ LLM System:
API Key: β
Configured
Success Rate: 87.5%
Avg Cost: $0.034/account
π§ RPA System:
BitBrowser: β
Running
Port: 35100
Success Rate: 92.1%
Avg Cost: $0.018/account
π Hybrid Controller:
Total Sessions: 156
LLM Preferred: 89 (57%)
RPA Preferred: 67 (43%)
Fallbacks: 12 (7.7%)# Run all tests
npm run test:integration
# Run specific test suites
npm run test:llm # LLM components only
npm run test:rpa # RPA components only
npm run test:hybrid # Hybrid controller only
# Run with debug output
npm run kilocode:test -- --debugTests generate detailed HTML and JSON reports in ./test-results/:
integration-test-report.html- Human-readable reportintegration-test-report.json- Machine-readable data- Individual session logs and error reports
# Check API key configuration
npm run kilocode:status
# Test with debug mode
npm run kilocode:llm -- --debug
# Configure API key
npm run kilocode:config -- --openai-key "your-key"# Check BitBrowser status
npm run check-bitbrowser
# Test BitBrowser integration
npm run bitbrowser:test
# Restart BitBrowser and try again# Check system health
npm run kilocode:status
# Run performance tests
npm run kilocode:test -- --performance
# Monitor via dashboard
npm run kilocode:dashboardEnable debug mode for detailed logging:
# Any command with debug
npm run kilocode:hybrid -- --debug
npm run kilocode:test -- --debug- Rate Limiting: Configured to stay within API limits
- Cost Management: Token counting and cost tracking
- Intelligent Caching: Reuse responses where appropriate
- Fallback Data: Backup generation when API fails
- Resource Blocking: Block images/CSS for faster execution
- Profile Reuse: Efficiently manage BitBrowser profiles
- Parallel Execution: Multiple profiles simultaneously
- Error Recovery: Automatic retry and recovery mechanisms
- Success Rate Learning: Adapts strategy based on performance
- Complexity Analysis: Routes complex tasks to LLM
- Cost Optimization: Balances quality vs cost
- Failover Logic: Smart fallback between approaches
- Environment variable configuration
- Rate limiting and quota management
- Error handling without exposing sensitive data
- Secure credential storage
- Antidetect profile management
- Proxy rotation support
- Session isolation
- Fingerprint randomization
- Comprehensive logging without sensitive data
- Error tracking and monitoring
- Health checks and alerting
- Graceful degradation
gmail kilocode creator bybitbot/
βββ π€ LLM Components
β βββ kilocode-llm-api.js # LLM API integration
β βββ kilocode-form-engine.js # Intelligent form filling
βββ π§ RPA Components
β βββ real-browser-automation.js # Enhanced browser automation
β βββ bitbrowser-profile-manager.js # Profile management
β βββ automa-*.json # Workflow definitions
βββ π― Hybrid System
β βββ hybrid-automation-controller.js # Main controller
β βββ unified-dashboard.js # Web dashboard
βββ π§ͺ Testing
β βββ comprehensive-integration-tests.js # Test suite
βββ π Main Interface
β βββ kilocode-main.js # CLI interface
β βββ package.json # Scripts and dependencies
βββ π Reports & Logs
βββ test-results/ # Test reports
βββ logs/ # Execution logs
βββ reports/ # Session reports
- Fork the repository
- Create a feature branch (
git checkout -b feature/amazing-feature) - Commit your changes (
git commit -m 'Add amazing feature') - Push to the branch (
git push origin feature/amazing-feature) - Open a Pull Request
# Install development dependencies
npm install
# Run tests
npm run test:integration
# Start dashboard in development mode
npm run kilocode:dashboard -- --debug
# Check code health
npm run kilocode:statusThis project is licensed under the MIT License - see the LICENSE file for details.
This software is for educational and research purposes only. Users are responsible for complying with all applicable terms of service and local laws. The authors assume no responsibility for misuse of this software.
- OpenAI & Anthropic for LLM APIs
- BitBrowser for antidetect browser technology
- Playwright for browser automation capabilities
- Automa for workflow inspiration
- Node.js community for excellent tooling
Made with β€οΈ by KiloCode Creator
Bridging AI intelligence with traditional automation