review-workflow（审查工作流程）

Review Workflow

Overview

review-workflow provides complete quality assurance by orchestrating three complementary review skills into a comprehensive validation workflow.

Purpose: Complete quality assurance from quality gating through comprehensive review to functional testing

Component Skills:

1. skill-validator - Pass/fail quality gate (minimum standards)
2. review-multi - Comprehensive multi-dimensional review (1-5 scoring)
3. skill-tester - Functional testing in real scenarios

Workflow Pattern: Sequential (validator → review → testing)

Result: Complete quality assessment with deployment decision, quality scores, and functional validation

When to Use

• Pre-deployment quality assurance (complete validation before release)
• Comprehensive quality checks (all aspects validated)
• Certification process (validate skill meets all standards)
• Major update validation (ensure quality maintained after changes)
• Quality audit (periodic comprehensive checks)

Review Workflow

Step 1: Quality Gate Validation (skill-validator)

Purpose: Fast pass/fail check - ensures minimum standards met before deeper review

Component Skill: skill-validator

Process:

1. Run skill-validator on skill
2. Check all 4 validation operations:

- Structure validation - Content validation - Pattern validation - Production readiness

1. Result: PASS or FAIL

Decision Point:

• PASS → Proceed to Step 2 (comprehensive review)
• FAIL → Stop, fix critical issues, re-run Step 1

Outputs:

• Validation report (pass/fail)
• Critical issues list (if failed)
• Go/no-go decision

Time Estimate: 30-45 minutes

Step 2: Comprehensive Review (review-multi)

Purpose: Multi-dimensional quality assessment with 1-5 scoring

Component Skill: review-multi

Process:

1. Run all 5 review operations:

- Operation 1: Structure Review (5-10 min, automated) - Operation 2: Content Review (15-30 min, manual) - Operation 3: Quality Review (20-40 min, mixed) - Operation 4: Usability Review (30-60 min, manual testing) - Operation 5: Integration Review (15-25 min, manual)

1. Calculate weighted overall score
2. Map to grade (A/B/C/D/F)
3. Assess production readiness
4. Generate improvement recommendations

Outputs:

• Overall score (1.0-5.0)
• Grade (A/B/C/D/F)
• Per-dimension scores
• Production readiness assessment
• Prioritized improvement recommendations

Time Estimate: 1.5-2.5 hours

Step 3: Functional Testing (skill-tester)

Purpose: Validate skill works correctly in real scenarios

Component Skill: skill-tester

Process:

1. Run testing operations:

- Scenario Testing (test in realistic use case) - Example Validation (verify examples execute) - Integration Testing (test with other skills if applicable) - Usability Testing (assess ease of use)

1. Document test results
2. Identify any functional issues

Outputs:

• Scenario test results (success/failure)
• Example validation results (all pass/some fail)
• Integration test results
• Usability assessment
• Functional issues list (if any)

Time Estimate: 45-90 minutes

Post-Workflow: Deployment Decision

After completing all 3 steps:

Aggregate Results

Quality Gate (Step 1):

• PASS/FAIL status
• Critical issues (if any)

Comprehensive Review (Step 2):

• Overall score: X.X/5.0
• Grade: A/B/C/D/F
• Production readiness assessment
• Improvement recommendations

Functional Testing (Step 3):

• Scenario tests: Pass/Fail
• Examples: All working/Some broken
• Integration: Working/Issues
• Usability: Good/Needs improvement

Make Deployment Decision

Deploy Immediately:

• Quality Gate: PASS
• Overall Score: ≥4.5 (Grade A)
• Functional Tests: All passing
• Action: Deploy to production

Deploy with Notes:

• Quality Gate: PASS
• Overall Score: 4.0-4.4 (Grade B+)
• Functional Tests: Passing
• Action: Deploy, note improvements for next iteration

Hold for Improvements:

• Quality Gate: PASS
• Overall Score: 3.5-3.9 (Grade B-)
• OR Functional Tests: Some failures
• Action: Fix identified issues, re-run workflow

Do Not Deploy:

• Quality Gate: FAIL
• OR Overall Score: <3.5 (Grade C-F)
• OR Functional Tests: Major failures
• Action: Significant rework required

Best Practices

1. Always Run Complete Workflow

Practice: Run all 3 steps before deployment (don't skip)

Rationale: Each step catches different issues - comprehensive coverage

2. Fix Critical Issues Immediately

Practice: If Step 1 fails, fix issues before proceeding

Rationale: Wastes time to do comprehensive review on skill failing basic validation

3. Document All Findings

Practice: Record results from each step for complete record

Rationale: Enables tracking improvements over time, learning from patterns

4. Apply Improvements Iteratively

Practice: Use Step 2 recommendations to plan next version

Rationale: Continuous improvement - each iteration gets better

5. Re-Run After Major Changes

Practice: Re-run workflow after applying significant improvements

Rationale: Validates improvements effective, no regressions introduced

Quick Reference

The 3-Step Review Workflow

Total Time: 3-4 hours for complete QA

Workflow Pattern

skill-validator (Quality Gate)
    ↓
  PASS? ──No──> Fix issues, retry
    ↓ Yes
review-multi (Comprehensive Review)
    ↓
Overall Score + Grade + Recommendations
    ↓
skill-tester (Functional Testing)
    ↓
Test Results + Usability Assessment
    ↓
Deployment Decision (Deploy/Hold/Rework)

Deployment Decision Matrix

Integration with Development

Build skill (development-workflow) →
Review workflow (this skill) →
Apply improvements (skill-updater) →
Re-validate (review-workflow) →
Deploy

review-workflow provides complete quality assurance through orchestrated validation, review, and testing.