proffesor-for-testing__agentic-qe/docs/qe-reports-3-6-3/code-complexity-analysis.md

Code Complexity Analysis Report - AQE v3.6.3

Date: 2026-02-11 Scope: /workspaces/agentic-qe-new/v3/src/ Analyzer: V3 QE Code Complexity Analyzer (claude-opus-4-6)

---

Executive Summary

Metric	Value	Status
Total Files	940	-
Total Lines	474,973	-
Average File Size	505 lines	WARNING
Median File Size	436 lines	OK
Files >500 lines	391 (41.6%)	CRITICAL
Files >1000 lines	84 (8.9%)	CRITICAL
Files >2000 lines	9 (1.0%)	CRITICAL
Methods >100 lines	184	CRITICAL
Methods >200 lines	24	CRITICAL
God Classes (>20 methods)	30	CRITICAL
Cross-module coupling links	559	HIGH
Kernel -> Domain imports	13	ARCHITECTURE VIOLATION

Key Findings

1. 41.6% of all files violate the 500-line project standard. This is the single most significant compliance failure. The project standard in CLAUDE.md states "Keep files under 500 lines" -- yet 391 of 940 files exceed this limit.

2. 9 files exceed 2000 lines, with the worst being quality-assessment/coordinator.ts at 2,426 lines (4.85x the limit).

3. 19 CRITICAL complexity files were identified with composite scores above 0.80, combining file size, method count, conditional density, error handling, and coupling.

4. God classes are pervasive. 30 classes have more than 20 methods each, with qe-agent-registry.ts having 98 methods in a single file.

5. 13 domain coordinators share identical mixin patterns (MinCut + Consensus + Governance), suggesting boilerplate that could be abstracted into a base class.

6. Kernel imports from Domains -- kernel/kernel.ts imports all 13 domain plugins directly, violating the dependency inversion principle.

---

File Size Distribution

     0-100:  119 files  #######################################
   101-200:  106 files  ###################################
   201-300:   81 files  ###########################
   301-500:  243 files  ################################################################################
   501-750:  194 files  ################################################################
  751-1000:  113 files  #####################################
 1001-1500:   58 files  ###################
 1501-2000:   17 files  #####
     2001+:    9 files  ###

Compliance: 549 files (58.4%) are within the 500-line limit. 391 files (41.6%) are non-compliant.

---

Top 20 Most Complex Files (Sorted by Composite Score)

Composite score formula: size(0.25) + methods(0.25) + conditionals(0.25) + error_handling(0.15) + coupling(0.10)

Rank	File	Lines	Methods	Ifs	Try/Catch	Imports	Score	Rating
1	domains/code-intelligence/coordinator.ts	2,156	82	82	27	27	1.00	CRITICAL
2	coordination/queen-coordinator.ts	2,202	90	117	20	29	1.00	CRITICAL
3	domains/quality-assessment/coordinator.ts	2,426	72	106	25	19	0.99	CRITICAL
4	domains/test-generation/coordinator.ts	1,845	67	70	16	19	0.94	CRITICAL
5	domains/visual-accessibility/services/accessibility-tester.ts	2,126	62	89	18	8	0.94	CRITICAL
6	domains/learning-optimization/coordinator.ts	2,094	64	74	12	17	0.94	CRITICAL
7	domains/security-compliance/services/security-auditor.ts	2,228	52	110	23	7	0.94	CRITICAL
8	domains/contract-testing/coordinator.ts	1,691	46	80	15	18	0.93	CRITICAL
9	cli/commands/learning.ts	1,726	52	86	22	13	0.93	CRITICAL
10	domains/chaos-resilience/coordinator.ts	1,889	55	74	13	16	0.93	CRITICAL
11	domains/visual-accessibility/coordinator.ts	1,825	60	69	16	16	0.92	CRITICAL
12	init/init-wizard.ts	2,113	57	66	19	10	0.91	CRITICAL
13	coordination/workflow-orchestrator.ts	2,219	61	82	9	4	0.86	CRITICAL
14	domains/requirements-validation/coordinator.ts	1,452	53	65	13	17	0.85	CRITICAL
15	domains/test-generation/services/pattern-matcher.ts	1,725	60	81	9	8	0.85	CRITICAL
16	kernel/unified-memory.ts	2,070	55	100	6	7	0.85	CRITICAL
17	domains/code-intelligence/services/c4-model/index.ts	1,603	50	83	11	4	0.83	CRITICAL
18	domains/security-compliance/coordinator.ts	1,358	53	61	12	18	0.82	CRITICAL
19	domains/contract-testing/services/contract-validator.ts	1,824	41	169	10	5	0.81	CRITICAL
20	domains/chaos-resilience/services/chaos-engineer.ts	1,176	62	65	15	9	0.80	HIGH

---

Files Exceeding 500-Line Limit

Total violations: 391 files (41.6% of codebase)

Files >2000 Lines (CRITICAL -- 4x+ Over Limit)

File	Lines	Over By
domains/quality-assessment/coordinator.ts	2,426	+1,926 (385%)
domains/security-compliance/services/security-auditor.ts	2,228	+1,728 (346%)
coordination/workflow-orchestrator.ts	2,219	+1,719 (344%)
coordination/queen-coordinator.ts	2,202	+1,702 (340%)
domains/code-intelligence/coordinator.ts	2,156	+1,656 (331%)
domains/visual-accessibility/services/accessibility-tester.ts	2,126	+1,626 (325%)
init/init-wizard.ts	2,113	+1,613 (323%)
domains/learning-optimization/coordinator.ts	2,094	+1,594 (319%)
kernel/unified-memory.ts	2,070	+1,570 (314%)

Distribution of Violations by Module

Module	Files	Total Lines	Files >500	Avg Lines	Severity
domains	209	127,615	108	610	CRITICAL
integrations	115	58,126	49	505	HIGH
coordination	88	49,428	37	561	CRITICAL
adapters	74	42,244	49	570	CRITICAL
mcp	89	36,343	29	408	HIGH
shared	61	25,062	25	410	MEDIUM
learning	28	21,053	22	751	CRITICAL
cli	50	21,008	14	420	HIGH
governance	16	13,159	12	822	CRITICAL
init	35	10,771	5	307	LOW
strange-loop	19	8,027	6	422	MEDIUM
kernel	14	5,624	1	401	LOW
routing	9	4,594	4	510	MEDIUM
planning	5	3,899	3	779	HIGH

Worst offenders by average file size: governance (822 avg), learning (751 avg), planning (779 avg), domains (610 avg).

Specifically Flagged: CLI Index (842 lines)

/workspaces/agentic-qe-new/v3/src/cli/index.ts is 845 lines, exceeding the 500-line limit by 69%. This file registers CLI commands using Commander.js and should be split into separate command registration modules.

---

God Classes (>20 Methods Per Class)

Methods	Lines	File
98	1,273	routing/qe-agent-registry.ts
93	1,075	integrations/coherence/types.ts
90	2,202	coordination/queen-coordinator.ts
82	2,156	domains/code-intelligence/coordinator.ts
72	2,426	domains/quality-assessment/coordinator.ts
68	1,401	domains/test-execution/services/user-flow-generator.ts
67	1,845	domains/test-generation/coordinator.ts
66	545	memory/crdt/types.ts
64	912	domains/coverage-analysis/services/gap-detector.ts
64	713	memory/crdt/crdt-store.ts
64	2,094	domains/learning-optimization/coordinator.ts
64	1,478	coordination/mincut/dream-integration.ts
63	1,212	integrations/ruvector/sona-persistence.ts
62	2,126	domains/visual-accessibility/services/accessibility-tester.ts
62	1,176	domains/chaos-resilience/services/chaos-engineer.ts
61	2,219	coordination/workflow-orchestrator.ts
60	1,825	domains/visual-accessibility/coordinator.ts
60	1,725	domains/test-generation/services/pattern-matcher.ts
58	1,556	coordination/mincut/neural-goap.ts
58	1,383	adapters/ag-ui/event-adapter.ts

Pattern: Nearly all domain coordinators are God classes. The coordinator pattern has accumulated too many responsibilities over time.

---

Longest Methods (Cognitive Complexity Hotspots)

Methods exceeding 100 lines are extremely difficult to test and understand.

Lines	Method	File	Issue
1,008	createAgentProfile (data block)	routing/qe-agent-registry.ts:60	Giant data literal -- not a method but a massive array declaration
622	registerBuiltInWorkflows	coordination/workflow-orchestrator.ts:1492	15+ workflow definitions crammed into one method
609	parseGraphQLField	contract-testing/services/contract-validator.ts:1216	Recursive parser with deep nesting
515	calculateComplexity	defect-intelligence/services/defect-predictor.ts:651	Multi-branch AST analysis with fallback heuristics
469	estimateComplexity	test-generation/services/pattern-matcher.ts:853	Large switch/case with string pattern matching
460	registerAllTools	mcp/protocol-server.ts:425	Tool registration -- 30+ tools inline
422	generateMockValue	test-generation/services/pattern-matcher.ts:420	Type-to-mock-value mapping with many branches
366	apply	mcp/tools/qx-analysis/heuristics-engine.ts:55	Rule engine with accumulated conditionals
327	generateTestBlock	test-execution/services/user-flow-generator.ts:1048	Template generation with many conditionals
299	getTestTemplates	requirements-validation/.../test-idea-generator.ts:474	Large data literal
281	performDASTScan	security-compliance/services/security-auditor.ts:1505	Complex security scan orchestration
265	getDemoSearchResult	mcp/tools/code-intelligence/analyze.ts:336	Hardcoded demo data
242	generateCLAUDEmdContent	init/init-wizard.ts:1521	Template string generation
233	generateTestValue	test-generation/generators/base-test-generator.ts:64	Type-to-value mapping
227	analyzeRequirements	requirements-validation/.../brutal-honesty-analyzer.ts:288	Multi-step analysis pipeline
218	initialize	init/init-wizard.ts:338	13-phase initialization sequence
217	execute	mcp/tools/visual-accessibility/index.ts:371	Multi-mode dispatch
206	getMetrics	adapters/a2a/discovery/metrics.ts:232	Metrics computation
199	performUnifiedAssertion	test-execution/services/e2e/assertion-handlers.ts:46	Multi-type assertion switch
186	calculateCognitiveComplexity	shared/metrics/code-metrics.ts:322	AST traversal with nesting tracking

Total methods >100 lines: 184 Total methods >200 lines: 24

---

Conditional Density Hotspots (Highest if/line Ratio)

The contract-validator.ts has the highest conditional density in the entire codebase: 169 if statements in 1,824 lines (9.3% of lines are conditionals). This indicates validation logic that should be decomposed into a rule-based or schema-driven validator.

File	Ifs	Lines	Density
contract-testing/services/contract-validator.ts	169	1,824	9.3%
coordination/queen-coordinator.ts	117	2,202	5.3%
security-compliance/services/security-auditor.ts	110	2,228	4.9%
domains/quality-assessment/coordinator.ts	106	2,426	4.4%
kernel/unified-memory.ts	100	2,070	4.8%
visual-accessibility/services/accessibility-tester.ts	89	2,126	4.2%
cli/commands/learning.ts	86	1,726	5.0%
code-intelligence/services/c4-model/index.ts	83	1,603	5.2%
governance/ab-benchmarking.ts	83	1,583	5.2%
coordination/workflow-orchestrator.ts	82	2,219	3.7%

---

Coupling Analysis Between Bounded Contexts

Cross-Module Import Counts

Source	Target	Import Count	Assessment
domains	shared	170	Expected (OK)
domains	kernel	59	Expected (OK)
domains	integrations	41	Acceptable
coordination	shared	35	Expected (OK)
integrations	shared	31	Expected (OK)
domains	coordination	27	WARNING -- Domains should not depend on coordination
mcp	domains	19	OK -- MCP dispatches to domains
learning	shared	18	Expected (OK)
coordination	kernel	17	Expected (OK)
mcp	kernel	14	OK
kernel	domains	13	ARCHITECTURE VIOLATION
learning	kernel	12	OK
mcp	learning	11	OK
mcp	coordination	8	OK
integrations	learning	8	Acceptable
coordination	domains	7	WARNING -- Creates circular dependency risk
shared	learning	7	ARCHITECTURE VIOLATION
integrations	kernel	6	OK

Architecture Violations

1. Kernel -> Domains (13 imports) -- CRITICAL

File: /workspaces/agentic-qe-new/v3/src/kernel/kernel.ts

The kernel directly imports all 13 domain plugin factories:

import { createTestGenerationPlugin } from '../domains/test-generation/plugin';
import { createTestExecutionPlugin } from '../domains/test-execution/plugin';
import { createCoverageAnalysisPlugin } from '../domains/coverage-analysis/plugin';
// ... 10 more domain imports

This violates dependency inversion. The kernel should define interfaces, and domains should register themselves through a plugin registry pattern.

2. Domains -> Coordination (27 imports) -- WARNING

Domain coordinators import from the coordination layer (mixins, consensus, event routing). While some of this is acceptable for mixin application, it creates tight coupling between the domain and coordination layers.

3. Coordination -> Domains (7 imports) -- WARNING

The coordination layer (particularly task-executor.ts) imports concrete domain services:

import { CoverageAnalyzerService } from '../domains/coverage-analysis';
import { SecurityScannerService } from '../domains/security-compliance';
import { createTestGeneratorService } from '../domains/test-generation';

This should use domain interfaces, not concrete implementations.

Inter-Domain Coupling

No direct inter-domain imports were detected (domain A does not import from domain B). This is a positive finding -- the bounded context separation between the 13 domains is clean at the import level.

---

Duplication Hotspots

1. Coordinator Mixin Boilerplate (All 13 Coordinators)

All 13 domain coordinators apply the same three mixins: MinCutAwareDomainMixin, ConsensusEnabledMixin, and GovernanceAwareMixin. Each coordinator contains duplicated:

• Mixin initialization code
• Health check aggregation from mixins
• Shutdown teardown for mixin resources
• Event subscription wiring

Estimated duplicated lines per coordinator: ~100-150 lines Total estimated duplication: ~1,300-1,950 lines

Recommendation: Create an AbstractDomainCoordinator base class that applies all standard mixins and provides the common lifecycle (initialize, shutdown, health check).

2. Repeated Initialization Patterns

All 13 coordinators implement async initialize(): Promise<void> with similar patterns:

• Initialize sub-services
• Subscribe to domain events
• Register with coordination layer
• Set up health monitoring

3. Tool Registration Patterns (MCP)

registerAllTools() in mcp/protocol-server.ts (460 lines) and similar patterns in handlers contain repetitive tool definition structures. This should use a declarative schema.

4. Test Output Parsers (Flaky Detector)

flaky-detector.ts contains three nearly identical parser methods:

• parseVitestJson() (lines 692-724)
• parseJestJson() (lines 729-762)
• parseTapOutput() (lines 767-797)

Each follows the same pattern of iterating test results and creating TestExecutionRecord objects. These could share a common parsing pipeline.

---

Refactoring Recommendations (Prioritized)

Priority 1: CRITICAL -- Immediate Action Required

R1. Extract Domain Coordinator Base Class

Files: All 13 coordinators (26,000+ lines total) Issue: Duplicated mixin application, lifecycle, and health check patterns Strategy: Create AbstractDomainCoordinator<TConfig> with standard mixin application Estimated reduction: ~1,500 lines total, each coordinator reduces by ~100-150 lines Testability improvement: Mixins testable once in base class, not 13x

R2. Split Queen Coordinator (2,202 lines, 90 methods)

File: coordination/queen-coordinator.ts Issue: Single class doing task routing, domain management, scaling, tracing, federation, hypothesis management, work stealing, and TinyDancer routing Strategy: Extract into focused classes:

• TaskRouter -- task assignment and routing
• DomainManager -- domain lifecycle and health
• ScalingController -- dynamic scaling decisions
• FederationCoordinator -- cross-instance federation
• HypothesisCoordinator -- competing hypotheses Estimated reduction: 2,202 -> 5 files of ~400-500 lines each Testability improvement: 4x (each class independently testable)

R3. Split Quality Assessment Coordinator (2,426 lines, 72 methods)

File: domains/quality-assessment/coordinator.ts Issue: Combines quality gate evaluation, deployment advice, complexity analysis, claim verification, consensus verification, RL integration, and SONA learning Strategy: Extract:

• QualityGateEngine -- gate evaluation and borderline detection
• ConsensusVerifier -- consensus-based verification
• RLOptimizer -- reinforcement learning integration
• ClaimVerificationService -- report claim checking Estimated reduction: 2,426 -> 4 files of ~500-600 lines

R4. Invert Kernel-Domain Dependency

File: kernel/kernel.ts Issue: Kernel directly imports 13 domain plugins (dependency inversion violation) Strategy: Implement plugin registry where domains self-register

// Instead of: import { createTestGenerationPlugin } from '../domains/...'
// Use: kernel.registerPlugin(plugin) called from each domain's index.ts

Impact: Eliminates architectural violation, enables lazy domain loading

Priority 2: HIGH -- Address Within Sprint

R5. Decompose Workflow Orchestrator (2,219 lines)

File: coordination/workflow-orchestrator.ts Issue: registerBuiltInWorkflows() is 622 lines of inline workflow definitions Strategy: Move workflow definitions to declarative YAML/JSON files, load dynamically Estimated reduction: 2,219 -> ~800 lines + separate workflow definition files

R6. Split Security Auditor (2,228 lines)

File: domains/security-compliance/services/security-auditor.ts Issue: Single class handling SAST, DAST, secret scanning, dependency auditing Strategy: Extract per-scan-type classes: SASTScanner, DASTScanner, SecretScanner, DependencyAuditor Estimated reduction: 2,228 -> 4 files of ~500-600 lines

R7. Refactor Contract Validator Conditionals (169 if-statements)

File: domains/contract-testing/services/contract-validator.ts Issue: Highest conditional density in codebase (9.3%) Strategy: Replace validation if-chains with schema-driven validation rules Estimated reduction: 169 ifs -> ~40 ifs + validation rule definitions

R8. Split Init Wizard (2,113 lines)

File: init/init-wizard.ts Issue: generateCLAUDEmdContent() is 242 lines, initialize() is 218 lines, configureHooks() is 193 lines Strategy: Extract CLAUDEmdGenerator, HooksConfigurator, and use phase-based initialization (already partially done) Estimated reduction: 2,113 -> ~800 + extracted generators

R9. Split Unified Memory (2,070 lines)

File: kernel/unified-memory.ts Issue: Manages KV store, vectors, HNSW, Q-values, GOAP, dreams, CRDT, hypergraph all in one class Strategy: Extract storage-specific adapters: KVStoreAdapter, VectorStoreAdapter, HNSWIndexAdapter Estimated reduction: 2,070 -> 4-5 files of ~400-500 lines

R10. Decompose registerAllTools (460 lines)

File: mcp/protocol-server.ts Issue: All 30+ MCP tools registered in a single method Strategy: Use tool auto-discovery: each tool file exports a definition that protocol-server discovers Estimated reduction: 460 -> ~50 lines (routing only) + tools self-register

Priority 3: MEDIUM -- Technical Debt Backlog

R11. Split CLI Index (845 lines)

File: cli/index.ts Issue: 845 lines registering Commander.js commands Strategy: Already partially done with cli/commands/ modules; move remaining inline command definitions to separate files Estimated reduction: 845 -> ~300 lines

R12. Consolidate Agent Registry Data (1,273 lines)

File: routing/qe-agent-registry.ts Issue: 98 agent profile definitions as inline data (1,008-line data block) Strategy: Move agent definitions to a JSON/YAML configuration file Estimated reduction: 1,273 -> ~300 lines + config file

R13. Extract Test Output Parsers

File: domains/test-execution/services/flaky-detector.ts Issue: Three near-identical test output parsers (Vitest, Jest, TAP) Strategy: Create TestOutputParser interface with framework-specific implementations Estimated reduction: ~100 lines (deduplication)

R14. Split Code Intelligence Coordinator (2,156 lines)

File: domains/code-intelligence/coordinator.ts Issue: 82 methods, 27 imports Strategy: Extract KnowledgeGraphCoordinator, C4ModelCoordinator, SemanticAnalysisCoordinator

R15. Split Pattern Matcher (1,725 lines)

File: domains/test-generation/services/pattern-matcher.ts Issue: estimateComplexity() is 469 lines, generateMockValue() is 422 lines Strategy: Extract ComplexityEstimator and MockValueGenerator as standalone services

---

Testability Assessment

Most Difficult to Test (Score <40/100)

File	Methods	Dependencies	Side Effects	Testability
queen-coordinator.ts	90	15+ services	Event bus, timers, scaling	Very Difficult
quality-assessment/coordinator.ts	72	RL, SONA, Consensus, MinCut	DB, events, ML models	Very Difficult
code-intelligence/coordinator.ts	82	KnowledgeGraph, C4, AST	File system, DB	Very Difficult
unified-memory.ts	55	SQLite, CRDT, HNSW	Database file I/O	Difficult
workflow-orchestrator.ts	61	All 13 domains	Events, timers	Very Difficult

Testing Effort Estimate

Complexity Level	Files	Estimated Hours	Tests Needed
CRITICAL (>0.80)	19	~380h	~1,900
HIGH (0.60-0.80)	21	~210h	~1,050
MEDIUM (0.40-0.60)	~100	~500h	~3,000
LOW (<0.40)	~251	~500h	~3,750
Total	391	~1,590h	~9,700

---

Maintainability Index Summary

Using a simplified maintainability model based on file size, complexity, coupling, and duplication:

Module	Files	Maintainability	Trend
kernel	14	68/100 MEDIUM	Stable
domains	209	42/100 LOW	Declining
coordination	88	45/100 LOW	Declining
adapters	74	50/100 MEDIUM	Stable
mcp	89	55/100 MEDIUM	Stable
learning	28	38/100 LOW	Declining
governance	16	40/100 LOW	Stable
integrations	115	48/100 LOW	Stable
Overall	940	46/100 LOW	Declining

---

Appendix: Complete List of Files >1500 Lines

Lines	File Path
2,426	v3/src/domains/quality-assessment/coordinator.ts
2,228	v3/src/domains/security-compliance/services/security-auditor.ts
2,219	v3/src/coordination/workflow-orchestrator.ts
2,202	v3/src/coordination/queen-coordinator.ts
2,156	v3/src/domains/code-intelligence/coordinator.ts
2,126	v3/src/domains/visual-accessibility/services/accessibility-tester.ts
2,113	v3/src/init/init-wizard.ts
2,094	v3/src/domains/learning-optimization/coordinator.ts
2,070	v3/src/kernel/unified-memory.ts
1,889	v3/src/domains/chaos-resilience/coordinator.ts
1,860	v3/src/domains/requirements-validation/qcsd-refinement-plugin.ts
1,845	v3/src/domains/test-generation/coordinator.ts
1,825	v3/src/domains/visual-accessibility/coordinator.ts
1,824	v3/src/domains/contract-testing/services/contract-validator.ts
1,730	v3/src/cli/completions/index.ts
1,726	v3/src/cli/commands/learning.ts
1,725	v3/src/domains/test-generation/services/pattern-matcher.ts
1,697	v3/src/domains/requirements-validation/qcsd-ideation-plugin.ts
1,691	v3/src/domains/contract-testing/coordinator.ts
1,637	v3/src/shared/llm/router/types.ts
1,603	v3/src/domains/code-intelligence/services/c4-model/index.ts
1,595	v3/src/coordination/mincut/time-crystal.ts
1,587	v3/src/coordination/protocols/security-audit.ts
1,583	v3/src/governance/ab-benchmarking.ts
1,566	v3/src/coordination/protocols/quality-gate.ts
1,556	v3/src/coordination/mincut/neural-goap.ts

---

Report generated by V3 QE Code Complexity Analyzer on 2026-02-11. All file paths are relative to /workspaces/agentic-qe-new/v3/src/ unless otherwise noted.