Review Task

Quality Review Orchestrator

Orchestrate comprehensive quality assessment by executing 5 specialized quality skills in sequence and synthesizing results into a unified quality gate decision.

Purpose

This skill coordinates execution of 5 specialized quality skills to provide comprehensive quality assessment:

1. risk-profile → Assess implementation risks using P×I methodology
2. test-design → Design comprehensive test strategy with P0/P1/P2 priorities
3. trace-requirements → Map AC → Implementation → Tests with gap analysis
4. nfr-assess → Assess non-functional requirements (security, performance, reliability, etc.)
5. quality-gate → Synthesize all assessments into final gate decision

When to Use This Skill

This skill should be used when: Comprehensive quality review of completed task (status "Review"), preparing for merge/deploy decision

Individual assessments: Invoke specific skills directly (risk-profile, test-design, trace-requirements, nfr-assess, quality-gate)

This skill should NOT be used when: Task in progress, reviewing drafts, quick spot checks

Architecture

Quality Review Orchestration Flow:

Step 0: Configuration & Verification
  ↓
Step 1: risk-profile → .claude/quality/assessments/{task-id}-risk-{date}.md
  ↓
Step 2: test-design → .claude/quality/assessments/{task-id}-test-design-{date}.md
  ↓
Step 3: trace-requirements → .claude/quality/assessments/{task-id}-trace-{date}.md
  ↓
Step 4: nfr-assess → .claude/quality/assessments/{task-id}-nfr-{date}.md
  ↓
Step 5: quality-gate → .claude/quality/gates/{task-id}-gate-{date}.{yaml,md}
  ↓
Step 6: Update Task File
  ↓
Step 7: Present Summary

Each skill builds on previous results. Execution order is critical.

File Modification Permissions

AUTHORIZED: Read task/implementation files, execute quality skills, generate reports, update task "Quality Review" section only

NOT AUTHORIZED: Modify task Objective/AC/Context/Tasks/Implementation, modify code, change status to "Done", bypass skills

Sequential Orchestration Execution

Step 0: Configuration and Verification

Purpose: Load configuration, verify task readiness, determine execution mode.

Actions:

1. Load configuration from .claude/config.yaml
2. Verify task file exists, status is "Review", implementation complete
3. Determine execution mode (Full/Individual/Resume)
4. Check existing assessments (for resume mode)

Halt if: Config missing, task not ready, status not "Review"

See: references/execution-modes-guide.md for detailed flow and references/templates.md for configuration format

Step 1: Execute Risk Profile Skill

Purpose: Assess implementation risks using P×I (Probability × Impact) methodology.

Actions:

1. Invoke risk-profile.md skill:

   Executing: .claude/skills/quality/risk-profile.md
   Task: {task-id}
   

2. Skill executes 7-step risk assessment process:

   • Identify risk areas (10-20 risks)
   • Score risks (P×I, scale 1-9)
   • Develop mitigation strategies
   • Prioritize test scenarios
   • Generate risk profile report

3. Capture results: risk count, critical/high risks, mitigation coverage, report path

Error Handling: If skill fails, ask user to fix and re-run, or skip and continue (impacts gate confidence).

See: references/templates.md for output format and captured data structure

Step 2: Execute Test Design Skill

Purpose: Design comprehensive test strategy with P0/P1/P2 priorities and mock strategies.

Actions:

1. Invoke test-design.md skill:

   Executing: .claude/skills/quality/test-design.md
   Task: {task-id}
   Risk Profile: {risk-file} (for risk-informed test prioritization)
   

2. Skill executes 7-step test design process:

   • Analyze test requirements per AC
   • Design test scenarios (Given-When-Then)
   • Develop mock strategies
   • Plan CI/CD integration
   • Generate test design document

3. Capture results: total tests, P0/P1/P2 counts, test levels, report path

See: references/templates.md for output format

Step 3: Execute Requirements Traceability Skill

Purpose: Map acceptance criteria → implementation → tests with gap analysis.

Actions:

1. Invoke trace-requirements.md skill:

   Executing: .claude/skills/quality/trace-requirements.md
   Task: {task-id}
   Risk Profile: {risk-file}
   Test Design: {test-file}
   

2. Skill executes 7-step traceability process:

   • Build forward traceability (AC → Implementation)
   • Build backward traceability (Tests → AC)
   • Identify coverage gaps
   • Create traceability matrix
   • Generate recommendations

3. Capture results: traceability score, implementation/test coverage, gaps, report path

Error Handling: Traceability is CRITICAL for quality gate. Must fix and re-run if fails.

See: references/templates.md for output format

Step 4: Execute NFR Assessment Skill

Purpose: Assess non-functional requirements across 6 categories.

Actions:

1. Invoke nfr-assess.md skill:

   Executing: .claude/skills/quality/nfr-assess.md
   Task: {task-id}
   Risk Profile: {risk-file}
   Traceability: {trace-file}
   Test Design: {test-file}
   

2. Skill executes 8-step NFR assessment:

   • Security assessment (validation, auth, encryption, vulnerabilities)
   • Performance assessment (response time, queries, caching)
   • Reliability assessment (error handling, logging, monitoring)
   • Maintainability assessment (code quality, docs, coverage)
   • Scalability assessment (stateless, indexing, async processing)
   • Usability assessment (API design, error messages, docs)

3. Capture results: overall NFR score, category scores (security/performance/reliability/maintainability/scalability/usability), critical gaps, report path

See: references/templates.md for output format

Step 5: Execute Quality Gate Skill

Purpose: Synthesize all assessments and make final PASS/CONCERNS/FAIL/WAIVED decision.

Actions:

1. Invoke quality-gate.md skill:

   Executing: .claude/skills/quality/quality-gate.md
   Task: {task-id}
   Risk Profile: {risk-file}
   Test Design: {test-file}
   Traceability: {trace-file}
   NFR Assessment: {nfr-file}
   

2. Skill executes 8-step gate synthesis:

   • Synthesize risk management dimension
   • Synthesize test coverage dimension
   • Synthesize traceability dimension
   • Synthesize NFR dimension
   • Synthesize implementation quality dimension
   • Synthesize compliance dimension
   • Calculate overall quality score
   • Make gate decision (PASS/CONCERNS/FAIL/WAIVED)

3. Capture results: gate decision (PASS/CONCERNS/FAIL/WAIVED), overall score, can proceed status, blockers, action items, report paths (YAML + MD)

Error Handling: Gate decision is CRITICAL. Must fix and re-run if fails.

See: references/templates.md for output format

Step 6: Update Task File with Quality Review Summary

Purpose: Update task file Quality Review section with synthesized summary.

Actions:

1. Generate quality review summary from gate decision
2. Update task file Quality Review section (preserve all other sections)

See: references/templates.md for task file update template

Step 7: Present Unified Quality Review Summary

Purpose: Present comprehensive summary synthesizing all 5 skill results with gate decision, dimension scores, findings, action items, and next steps.

Actions:

• Display gate decision and overall score
• Show quality dimension scores (risk, test, trace, NFR, implementation, compliance)
• List critical findings and action items (P0/P1)
• Present generated report links
• Collect user decision (accept/review/address/waive/rerun)

See: references/templates.md for complete summary template and user decision handling

Execution Complete

Quality review orchestration complete when:

• All 5 skills executed successfully (or skipped intentionally)
• Quality gate decision made
• Gate reports generated (YAML + Markdown)
• Task file Quality Review section updated
• Unified summary presented to user
• User decision collected and actioned

Best Practices

1. Run full review when possible - Most comprehensive assessment
2. Use individual skills during development - Faster feedback loops
3. Re-run specific skills when data changes - Don't redo entire review
4. Always start with gate report - Best overview, links to all details
5. Track action items in issue system - Don't lose follow-up work
6. Document waivers properly - Include justification, timeline, owner
7. Automate in CI/CD - Use gate YAML for automated checks

References

• references/execution-modes-guide.md - Execution flow for full/individual/resume modes
• references/error-handling-guide.md - Error scenarios, graceful degradation, fallback strategies
• references/orchestration-guide.md - Skill coordination and sequencing details
• references/synthesis-summary-guide.md - Summary generation and presentation
• references/templates.md - Output templates, task file updates, user decision handling