Unified Development Orchestrator with Predictive Uncertainty Modeling
세계 최초의 예측적 불확실성 모델링을 적용한 지능형 개발 자동화 플랫폼입니다.
- Phase-Aware Evaluation: 개발 단계별 맞춤 평가 (Ideation → Testing)
- Predictive Uncertainty: 24시간 불확실성 예측 모델링
- Quantum States: 5단계 양자 불확실성 상태 분류
- Auto-Mitigation: 실시간 완화 전략 생성 with ROI 계산
- ML Learning System: 패턴 인식 및 지속적 개선
| Metric | v1 | v3 | Improvement |
|---|---|---|---|
| Ideation Confidence | 9% | 80% | +888% |
| Phase Recognition | 0% | 100% | ∞ |
| Uncertainty Prediction | None | 24h | ∞ |
| Auto-Mitigation | None | Real-time | ∞ |
┌─────────────────────────────────────┐
│ UDO v2 (Orchestrator) │
│ - Phase-Aware Evaluation │
│ - Bayesian Confidence │
└────────────┬────────────────────────┘
│
┌────────────▼────────────────────────┐
│ Uncertainty Map v3 (Predictor) │
│ - Predictive Modeling │
│ - Quantum States │
│ - Auto-Mitigation │
└────────────┬────────────────────────┘
│
┌────────────▼────────────────────────┐
│ AI Collaboration Bridge (3-AI) │
│ - Claude + Codex + Gemini │
└─────────────────────────────────────┘
UDO-Development-Platform/
├── src/ # Core source code
│ ├── unified_development_orchestrator_v2.py
│ ├── uncertainty_map_v3.py
│ └── three_ai_collaboration_bridge.py
├── tests/ # Test suites
│ ├── test_udo_v3_integration.py
│ └── run_udo_phase1.py
├── docs/ # Documentation
│ ├── UDO_V3_INTEGRATION_REPORT.md
│ └── udo_v2_upgrade_report.md
├── data/ # Data & state files
│ ├── udo_learning_data.json
│ └── udo_state_phase1.json
└── README.md
- 현재 PR과 문서는 UDO v2 + Uncertainty Map v3 구성을 표준으로 사용합니다.
- 버전 1 대비 개선 지표와 선택 근거는
docs/pr_version_selection.md에서 요약합니다. - 향후 v3 전환 계획은
docs/UDO_V3_INTEGRATION_REPORT.md와FINAL_REPORT.md를 참조하세요. - Codex 리팩토링 8단계(로깅 정비~타입 검증)는 위 문서 4장에서 세부 내역과 테스트 결과를 확인할 수 있습니다.
# Clone repository
git clone https://github.com/[username]/UDO-Development-Platform.git
cd UDO-Development-Platform
# Install dependencies
pip install -r requirements.txtfrom src.unified_development_orchestrator_v2 import UnifiedDevelopmentOrchestratorV2, ProjectContext
# Initialize project
project = ProjectContext(
project_name="My-AI-Project",
goal="AI-powered solution",
team_size=5,
timeline_weeks=12,
budget=50000,
tech_stack=["Next.js", "FastAPI", "PostgreSQL"],
constraints=["3 months deadline"],
success_metrics=["MAU 1000+"],
current_phase="ideation",
files=[],
metadata={}
)
# Create UDO instance
udo = UnifiedDevelopmentOrchestratorV2(project)
# Start development cycle
plan = udo.start_development_cycle("Build AI code review platform")
# Execute if GO decision
if plan['decision'] == 'GO':
result = udo.execute_plan(plan)4-Tier Governance System 관리를 위한 CLI 도구입니다.
.\udo.bat status출력 예시:
🛡️ UDO Governance Status
=========================
Current Tier: Tier 1: Experiment
Description: 실험/학습
Compliance: 100%
🚀 Next Level: tier-2
Run 'udo upgrade-tier --to=tier-2' to upgrade.
.\udo.bat upgrade-tier --to=tier-2기능:
- 자동으로 필요한 파일/폴더 생성
- Tier 2:
config/schema.py,tests/__init__.py - Tier 3:
src/domain,src/application,src/infrastructure등
자세한 내용: 4-Tier Governance Guide
Use the improved startup script with automatic process management:
# Start backend server (automatically handles port conflicts)
python scripts/start_dev.py
# Or specify custom port
python scripts/start_dev.py --port 8080
# Fast start (skip process checks)
python scripts/start_dev.py --no-checkFeatures:
- ✅ Automatic duplicate process detection
- ✅ Graceful shutdown of existing processes
- ✅ Port conflict resolution
- ✅ PID file tracking
- ✅ Clean exit on Ctrl+C
Check running processes:
python scripts/cleanup_processes.py --check-onlyClean up all UDO processes:
python scripts/cleanup_processes.pyFrontend development server:
cd web-dashboard
npm run devIf you prefer manual control:
# Backend (Windows)
.venv\Scripts\python.exe -m uvicorn backend.main:app --reload --host 0.0.0.0 --port 8000
# Backend (Linux/Mac)
python -m uvicorn backend.main:app --reload --host 0.0.0.0 --port 8000Note: Manual startup doesn't include automatic process management. See docs/PORT_CONFLICT_ROOT_CAUSE_ANALYSIS.md for details.
To run the test suite, use the run_tests.py script in the root directory:
python run_tests.pyThis will discover and run all tests in the tests/ directory.
- Market validation
- Technical feasibility
- Innovation assessment
- Architecture quality
- Pattern adherence
- Scalability analysis
- Core feature coverage
- Development speed
- Validation readiness
- Code quality metrics
- Test coverage
- Performance benchmarks
- Coverage completeness
- Edge case handling
- Regression prevention
| State | Symbol | Range | Description |
|---|---|---|---|
| DETERMINISTIC | 🟢 | <10% | Fully predictable |
| PROBABILISTIC | 🔵 | 10-30% | Statistical confidence |
| QUANTUM | 🟠 | 30-60% | Multiple possibilities |
| CHAOTIC | 🔴 | 60-90% | High uncertainty |
| VOID | ⚫ | >90% | Unknown territory |
# Run integration tests
python tests/test_udo_v3_integration.py
# Test specific phase
python tests/run_udo_phase1.pyContributions are welcome! Please read our contributing guidelines.
MIT License - see LICENSE file for details
- ✅ Phase-Aware Evaluation System
- ✅ Predictive Uncertainty Modeling
- ✅ Quantum State Classification
- ✅ Auto-Mitigation Strategy Generation
- ✅ ML-based Learning System
- 🔄 Real AI Integration (30% - in progress)
- 🔄 Production Deployment (60% - testing)
- GitHub: [Your GitHub Profile]
- Email: [Your Email]
Version: 3.0.0 Status: Beta Testing Last Updated: 2025-11-17