ac-complexity-assessor（交流复杂度评估器）

AC Complexity Assessor

Assess complexity for effort estimation and planning.

Purpose

Analyzes features and projects to determine complexity levels, estimate effort, and select appropriate processing pipelines.

Quick Start

from scripts.complexity_assessor import ComplexityAssessor

assessor = ComplexityAssessor(project_dir)
assessment = await assessor.assess_project()
print(assessment.pipeline_type)  # SIMPLE/STANDARD/COMPLEX
print(assessment.estimated_hours)

Assessment Output

{
  "project_complexity": "STANDARD",
  "pipeline_type": "STANDARD",
  "metrics": {
    "total_features": 75,
    "dependency_depth": 4,
    "category_count": 8,
    "integration_points": 5,
    "technology_complexity": 3
  },
  "estimates": {
    "total_hours": 120,
    "estimated_cost_usd": 45.00,
    "estimated_sessions": 15,
    "features_per_session": 5
  },
  "feature_estimates": [
    {
      "id": "auth-001",
      "complexity": "low",
      "estimated_hours": 1.5,
      "factors": ["standard_pattern", "no_dependencies"]
    }
  ],
  "recommendations": [
    "Consider parallelizing UI features",
    "auth-003 may need more time due to OAuth"
  ]
}

Complexity Levels

SIMPLE (1-3 phases)

• < 20 features
• Shallow dependencies (depth < 2)
• Single technology
• No external integrations

STANDARD (7 phases)

• 20-100 features
• Moderate dependencies (depth 2-5)
• 2-3 technologies
• Limited integrations

COMPLEX (8+ phases)

• 100 features
• Deep dependencies (depth > 5)
• Multiple technologies
• Many integrations

Complexity Factors

Feature Complexity

• Implementation difficulty
• Test complexity
• Documentation needs
• Integration requirements

Project Complexity

• Total feature count
• Dependency graph depth
• Technology stack breadth
• External integrations

Risk Factors

• Unfamiliar technologies
• Complex algorithms
• Security requirements
• Performance constraints

Pipeline Selection

SIMPLE Pipeline:
  1. Plan → 2. Implement → 3. Verify

STANDARD Pipeline:
  1. Analyze → 2. Plan → 3. Test Design
  4. Implement → 5. Test → 6. Review → 7. Verify

COMPLEX Pipeline:
  1. Deep Analyze → 2. Architecture → 3. Plan
  4. Test Design → 5. Implement → 6. Test
  7. Integration → 8. Review → 9. Verify

Cost Estimation

cost = (features * avg_tokens_per_feature * cost_per_token)
     + (sessions * session_overhead)
     + (complexity_multiplier * base_cost)

API Reference

See scripts/complexity_assessor.py for full implementation.