| name | complexity-router |
| description | Estimate task complexity and route high-complexity tasks (>=80%) to deep architectural review - use before plan generation |
Complexity Router Skill
Estimates task complexity using a weighted factor algorithm and routes high-complexity tasks (≥80%) to deep architectural review (human review, ultrathinking, or a dedicated architecture skill if available).
When to Use
- Before plan generation in Phase 4 of aa-ma-plan-workflow
- To determine appropriate level of review for a task
- When deciding between standard vs enhanced validation
- To identify tasks requiring deep reasoning
Complexity Estimation Algorithm
Input Factors
| Factor | Weight | Description |
|---|
| Scope | 25% | Number of files/components affected |
| Architectural Impact | 25% | Level of system changes |
| Technical Risk | 20% | Unknowns and technical uncertainty |
| Dependencies | 15% | External service/library count |
| Uncertainty | 15% | Requirements ambiguity |
Factor Scoring (0-100 each)
Scope (Files Affected)
1-3 files: 0-20
4-10 files: 21-40
11-20 files: 41-60
21-50 files: 61-80
50+ files: 81-100
Architectural Impact
No architecture changes: 0-20
Single component changes: 21-40
Multiple component changes: 41-60
Cross-service changes: 61-80
System-wide changes: 81-100
Technical Risk
Well-understood domain: 0-20
Minor unknowns: 21-40
Moderate unknowns: 41-60
Significant unknowns: 61-80
High uncertainty/new tech: 81-100
Dependencies
No external dependencies: 0-20
1-2 dependencies: 21-40
3-5 dependencies: 41-60
6-10 dependencies: 61-80
10+ dependencies: 81-100
Uncertainty
Clear requirements: 0-20
Minor ambiguity: 21-40
Moderate ambiguity: 41-60
Significant ambiguity: 61-80
High ambiguity: 81-100
Calculation
Complexity = (Scope × 0.25) + (ArchImpact × 0.25) +
(TechRisk × 0.20) + (Deps × 0.15) +
(Uncertainty × 0.15)
Routing Thresholds
| Complexity | Category | Routing Action |
|---|
| 0-29% | Low | Standard workflow, minimal review |
| 30-59% | Medium | Standard workflow with validation |
| 60-79% | High | Enhanced validation, flag in plan |
| 80-100% | Critical | Deep architectural review (human or dedicated skill) |
Auto-Trigger Indicators
The following automatically set complexity ≥80%:
Architectural Triggers
Risk Triggers
Scale Triggers
Procedure
Step 1: Gather Inputs
From Phase 2-3 outputs:
- Design concept and scope
- Research findings
- Impact assessment results
- Identified risks
Step 2: Score Each Factor
Rate each factor on 0-100 scale:
Scope: [__]/100
Arch Impact: [__]/100
Tech Risk: [__]/100
Dependencies: [__]/100
Uncertainty: [__]/100
Step 3: Check Auto-Triggers
Scan for automatic high-complexity indicators:
Auto-Triggers Found:
- [ ] [Trigger 1]
- [ ] [Trigger 2]
If any trigger found: Complexity = max(calculated, 80)
Step 4: Calculate and Route
Calculated Complexity: [X]%
Auto-Trigger Override: [Yes/No]
Final Complexity: [X]%
Routing Decision: [Standard / Enhanced / Senior-Architect]
Step 5: Generate Report
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
COMPLEXITY ASSESSMENT
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Task: [task-name]
Date: [YYYY-MM-DD]
FACTOR SCORES:
Scope: [XX]/100 (weight: 25%)
Architectural: [XX]/100 (weight: 25%)
Technical Risk: [XX]/100 (weight: 20%)
Dependencies: [XX]/100 (weight: 15%)
Uncertainty: [XX]/100 (weight: 15%)
──────────────────────────────────
WEIGHTED TOTAL: [XX]%
AUTO-TRIGGERS:
[✓/✗] [Trigger description]
[✓/✗] [Trigger description]
FINAL COMPLEXITY: [XX]% ([Low/Medium/High/Critical])
ROUTING DECISION: [Action]
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Integration with aa-ma-plan-workflow
This skill is invoked at the start of Phase 4:
- Complexity is assessed using this skill
- If complexity ≥80%, deep architectural review is triggered (human review, ultrathinking, or a dedicated architecture skill)
- Routing decision recorded in provenance.log
- Plan generation proceeds with appropriate rigor
Routing Actions
Low Complexity (0-29%)
- Standard 5-phase workflow
- Basic validation gates
- No additional review
Medium Complexity (30-59%)
- Standard 5-phase workflow
- Full validation gates
- Quality score must be ≥70%
High Complexity (60-79%)
- Full validation gates
- Quality score must be ≥75%
- High-complexity steps flagged
- Recommend milestone-by-milestone execution
Critical Complexity (80-100%)
- MUST trigger deep architectural review (human review, ultrathinking, or a dedicated architecture skill if available)
- Quality score must be ≥80%
- All high-complexity steps require deep reasoning
- Recommend human review before execution
Checklist
When using this skill:
