proffesor-for-testing__agentic-qe/docs/cleanup/handler-integration-analysis.md

MCP Handler Integration Analysis

Date: 2026-01-02 Purpose: Determine which unregistered MCP handlers should be integrated before cleanup

Executive Summary

After analyzing 14 unregistered MCP handlers against the 60+ registered tools, I recommend:

Category	Integrate	Delete	Decision Needed
Chaos Handlers	3	0	0
Integration Handlers	2	1	0
Filtered Handlers	3	3	0
TOTAL	8	4	0

---

1. Chaos Handlers - INTEGRATE ALL (3 files)

Status: NO equivalent registered tools exist. These add entirely new capabilities.

Handler	Capability	User Value	Recommendation
chaos-inject-latency.ts	Network latency injection	Test timeout handling, slow network resilience	INTEGRATE
chaos-inject-failure.ts	Failure injection (HTTP errors, connection refused)	Test error handling, circuit breakers	INTEGRATE
chaos-resilience-test.ts	Comprehensive resilience testing with templates	Full chaos engineering capability	INTEGRATE

Key Features:

• Pre-built templates: network-partition, high-latency, cascading-failure
• Blast radius control for safe testing
• Auto-rollback capability
• Resilience score calculation (0-100)
• Actionable recommendations

Integration Effort: Medium

• Add 3 tool definitions to tools.ts
• Register in server.ts
• Add to TOOL_NAMES enum
• Estimated: ~2 hours

Business Value: HIGH

• Chaos engineering is critical for microservice reliability
• No existing chaos testing MCP tools
• Aligns with QE agents like qe-chaos-engineer

---

2. Integration Handlers - INTEGRATE 2, DELETE 1

2.1 dependency-check.ts - INTEGRATE

Status: UNIQUE - No equivalent registered tool

Capability: Service dependency health checking with:

• Parallel/sequential health checks
• Retry with exponential backoff
• Critical service identification
• Health score calculation (0-100)
• Detailed service metrics (connections, memory, CPU)

User Value: Essential for integration testing - verify all services are healthy before running tests.

Integration Effort: Low (~1 hour)

2.2 integration-test-orchestrate.ts - INTEGRATE

Status: UNIQUE - No equivalent registered tool

Capability: Multi-service integration test orchestration:

• Sequential/parallel test execution
• Retry with backoff for flaky tests
• Test data generation
• Environment-aware (dev/staging/prod warnings)
• Execution mode control

User Value: Enables coordinated integration testing across microservices.

Integration Effort: Low (~1 hour)

2.3 contract-validate.ts - DELETE (Overlaps)

Status: DUPLICATES existing functionality

Registered Equivalents:

• qe_api_contract_validate - Same OpenAPI validation
• qe_api_contract_breaking_changes - Breaking change detection
• qe_api_contract_versioning - Version compatibility

The unregistered handler adds:

• GraphQL schema validation
• Message queue contract validation
• gRPC contract validation

Recommendation: DELETE and consider adding GraphQL/MQ/gRPC support to existing registered tools if needed.

---

3. Filtered Handlers - INTEGRATE 3, DELETE 3

Token Optimization Value

The filtered handlers provide 95-99% token reduction, but we need to consider overlap with existing tools.

3.1 INTEGRATE (Unique Value)

Handler	Token Reduction	Registered Equivalent	Recommendation
test-executor-filtered.ts	97.3%	None	INTEGRATE
performance-tester-filtered.ts	98.3%	None	INTEGRATE
quality-assessor-filtered.ts	97.5%	None	INTEGRATE

Rationale: These provide unique capabilities with massive token savings.

3.2 DELETE (Overlap with Registered Tools)

Handler	Token Reduction	Registered Equivalent	Recommendation
coverage-analyzer-filtered.ts	99%	coverage_analyze_sublinear, coverage_gaps_detect, coverage_analyze_stream	DELETE
flaky-detector-filtered.ts	98.5%	flaky_detect_statistical, flaky_analyze_patterns, flaky_stabilize_auto	DELETE
security-scanner-filtered.ts	97.2%	security_scan_comprehensive, qe_security_scan_comprehensive	DELETE

Rationale: Registered tools already provide similar functionality. Adding filtered versions would create confusion about which tool to use.

Alternative: Consider adding tokenOptimized: true parameter to existing registered tools.

---

4. Type Files

File	Status	Recommendation
src/mcp/types/chaos.ts	Required by chaos handlers	KEEP (if integrating chaos)
src/mcp/types/integration.ts	Required by integration handlers	KEEP (if integrating integration)

---

5. Integration Priority Order

Phase 1: High Value, Low Effort (Week 1)

1. Chaos Handlers (3 files) - New capability, no overlap
2. Integration Handlers (2 files) - Unique value, low effort

Phase 2: Token Optimization (Week 2)

3. Unique Filtered Handlers (3 files) - High token savings

Phase 3: Cleanup

4. Delete overlapping handlers (4 files)
5. Delete type files if handlers deleted

---

6. Implementation Checklist

To Integrate (8 handlers):

✅ src/mcp/handlers/chaos/chaos-inject-latency.ts
✅ src/mcp/handlers/chaos/chaos-inject-failure.ts
✅ src/mcp/handlers/chaos/chaos-resilience-test.ts
✅ src/mcp/handlers/integration/dependency-check.ts
✅ src/mcp/handlers/integration/integration-test-orchestrate.ts
✅ src/mcp/handlers/filtered/test-executor-filtered.ts
✅ src/mcp/handlers/filtered/performance-tester-filtered.ts
✅ src/mcp/handlers/filtered/quality-assessor-filtered.ts

To Delete (4 handlers):

❌ src/mcp/handlers/integration/contract-validate.ts
❌ src/mcp/handlers/filtered/coverage-analyzer-filtered.ts
❌ src/mcp/handlers/filtered/flaky-detector-filtered.ts
❌ src/mcp/handlers/filtered/security-scanner-filtered.ts

---

7. Estimated Token Savings After Integration

If integrating the 3 unique filtered handlers at scale (1,000 ops/day):

Handler	Ops/Day	Token Reduction	Annual Savings
test-executor-filtered	500	97.3%	~$15,000
performance-tester-filtered	300	98.3%	~$20,000
quality-assessor-filtered	200	97.5%	~$8,000
TOTAL	1,000	98%+	~$43,000/year

Estimates based on Claude API pricing at $0.003/1K tokens

---

Decision Required

Before proceeding with cleanup, please confirm:

1. Integrate chaos handlers? (Adds chaos engineering capability)
2. Integrate dependency-check and integration-test-orchestrate? (Adds integration testing capability)
3. Integrate unique filtered handlers? (Adds token optimization)
4. Delete overlapping handlers? (Remove redundant code)

---

Generated by comprehensive handler analysis