Claude Code Plugins

Community-maintained marketplace

Feedback
0
0

Validates Testing Phase completion before advancing to Finalization Phase. Checks test execution results, coverage targets, test quality, and environment health. Ensures 100% test pass rate required by quality gates.

Install Skill

1Download skill
2Enable skills in Claude

Open claude.ai/settings/capabilities and find the "Skills" section

3Upload to Claude

Click "Upload skill" and select the downloaded ZIP file

Note: Please verify skill by going through its instructions before using it.

SKILL.md

name phase-4-validator
description Validates Testing Phase completion before advancing to Finalization Phase. Checks test execution results, coverage targets, test quality, and environment health. Ensures 100% test pass rate required by quality gates.

Phase 4 (Testing) Validation Skill

Purpose: Automate validation of Testing Phase before advancing to Finalization Phase.

When to Use: When orchestrator needs to verify all tests are passing and ready for deployment.

Critical Rule: 100% Test Pass Rate Required

ALL work types (Feature/Bug/Hotfix/Docs/Refactor) require 100% test pass rate.

No exceptions. One failing test = Phase 4 fails.

How to Use This Skill

Executable Script: execute.sh

# Basic usage (no parameters required)
bash .claude/skills/phase-4-validator/execute.sh

# Show help
bash .claude/skills/phase-4-validator/execute.sh --help

Parameters:

  • No parameters required - validates current Docker environment

Script validates:

  • Environment health (Docker containers, database, API, web services)
  • Test execution results (100% pass rate REQUIRED for all tests)
  • Test coverage (API ≥80%, React ≥70%)
  • Test quality (no flaky tests, execution time, independence, cleanup)
  • Documentation (TEST_CATALOG, test reports, known issues, metrics)

Exit codes:

  • 0: All tests passing at 100%, ready for Finalization Phase
  • 1: Tests failing OR environment unhealthy OR quality gates not met

CRITICAL: This validator BLOCKS workflow if ANY test fails. 100% pass rate is mandatory.

Quality Gate Checklist (100% Test Pass Rate Required)

Environment Health (15 points)

  • All Docker containers running (5 points)
  • Database healthy and seeded (5 points)
  • API service responding (3 points)
  • Web service responding (2 points)

Test Execution (40 points - MUST BE 40/40)

  • Unit tests: 100% passing (15 points)
  • Integration tests: 100% passing (15 points)
  • E2E tests: 100% passing (10 points)

Test Coverage (15 points)

  • API coverage ≥ 80% (5 points)
  • React coverage ≥ 70% (5 points)
  • Critical paths covered (5 points)

Test Quality (15 points)

  • No flaky tests (5 points)
  • Test execution time acceptable (3 points)
  • Tests are independent (4 points)
  • Test data cleanup working (3 points)

Documentation (15 points)

  • TEST_CATALOG updated (5 points)
  • Test results documented (5 points)
  • Known issues documented (3 points)
  • Performance metrics recorded (2 points)

Usage Examples

From Orchestrator

Use the phase-4-validator skill to verify all tests are passing

Manual Validation

# Run complete validation
bash .claude/skills/phase-4-validator.md

Common Issues

Issue: Environment Not Healthy

Solution: Run environment checks first

./dev.sh
# Wait for all containers to be healthy
docker ps

Issue: Failing Tests

Solution: Phase 4 CANNOT advance with failing tests

  • Loop back to react-developer for UI test failures
  • Loop back to backend-developer for API test failures
  • Loop back to test-developer for test logic issues
  • test-executor handles environment issues only

Issue: Low Coverage

Warning but not blocker: Coverage below targets is a warning but doesn't block if all tests pass.

However: Should document in testing-notes.md as known limitation.

Issue: Flaky Tests

Solution: Flaky tests = unreliable validation

  • Must be fixed before advancing
  • Document patterns causing flakiness
  • Add to lessons learned

Zero Tolerance Policy

Testing Phase has ZERO TOLERANCE for failing tests.

Even a single failing test means:

  • ❌ Phase 4 validation FAILS
  • ❌ Cannot advance to Phase 5
  • ❌ Must loop back to implementation

Why: Deploying with failing tests risks production issues.

Output Format

{
  "phase": "testing",
  "status": "pass|fail",
  "score": 95,
  "maxScore": 100,
  "percentage": 95,
  "testResults": {
    "unit": {
      "total": 45,
      "passed": 45,
      "failed": 0,
      "passRate": 100
    },
    "integration": {
      "total": 12,
      "passed": 12,
      "failed": 0,
      "passRate": 100
    },
    "e2e": {
      "total": 8,
      "passed": 8,
      "failed": 0,
      "passRate": 100
    }
  },
  "coverage": {
    "api": 85,
    "web": 72
  },
  "environment": {
    "docker": "healthy",
    "database": "healthy",
    "api": "responding",
    "web": "responding"
  },
  "quality": {
    "flakyTests": 0,
    "executionTime": 125.3,
    "independent": true,
    "cleanupWorking": true
  },
  "readyForNextPhase": true
}

Integration with Quality Gates

ALL work types require 100% test pass rate:

  • Feature: 100% pass (40/40 test points) + coverage/quality
  • Bug Fix: 100% pass (40/40 test points) + coverage/quality
  • Hotfix: 100% pass (40/40 test points) + minimal other requirements
  • Refactoring: 100% pass (40/40 test points) + no regressions

Progressive Disclosure

Initial Context: Show pass/fail status only On Request: Show detailed test results and coverage On Failure: Show specific failing tests with logs On Pass: Show concise summary + quality metrics


Remember: Phase 4 is the quality gate. If tests aren't 100% passing, nothing advances. This protects production from bugs.