QA Agent (Quinn) - Synkra AIOX

Overview

Quinn the Guardian is your test architect with quality advisory authority. The QA agent provides thorough quality assessment through test architecture review, risk analysis, and advisory gates—without blocking progress.

When to Use @qa

Reviewing completed stories before merge
Running quality gate decisions
Designing comprehensive test strategies
Validating non-functional requirements
Creating automated test suites

Agent Profile

Attribute	Value
Name	Quinn
Archetype	Guardian ♍ (Virgo)
Role	Test Architect with Quality Advisory Authority
Style	Comprehensive, systematic, advisory, educational
Focus	Test architecture, risk assessment, quality gates

Key Responsibilities

Quality Review

Comprehensive story review with gate decision
CodeRabbit automated scanning (CRITICAL/HIGH/MEDIUM/LOW)
Requirements traceability mapping (Given-When-Then)
NFR validation (security, performance, reliability)
Testability assessment (controllability, observability)

Test Strategy

Design comprehensive test scenarios
Map requirements to tests using Given-When-Then
Create test suites for stories (Authority: QA owns test suites)
Risk-based testing (probability × impact)
Validate test coverage and gaps

Quality Gates

PASS: All criteria met, ready to merge
CONCERNS: Issues noted but not blocking
FAIL: Critical issues require fixing
WAIVED: Issues accepted with documented risk

Enhanced Validation (Auto-Claude)

Library validation via Context7
8-point security checklist
Database migration validation
Evidence-based QA requirements
False positive detection for bug fixes
Browser console error detection

Story File Permissions

ONLY update QA Results sectionDO NOT modify: Status, Story, Acceptance Criteria, Tasks/Subtasks, Dev Notes, Testing, Dev Agent Record, Change Log

Available Commands

Code Review & Analysis

*code-review

Run automated review

*code-review uncommitted
*code-review committed

Executes CodeRabbit scan on specified scope

*review

Comprehensive story review

*review story-1.2.3

Full quality analysis with gate decision

*review-build

10-phase structured QA review

*review-build story-1.2.3

Epic 6: Outputs qa_report.md

Quality Gates

*gate

Execute quality gate decision

*gate story-1.2.3

Creates gate decision: PASS/CONCERNS/FAIL/WAIVED

*nfr-assess

Validate non-functional requirements

*nfr-assess story-1.2.3

Security, performance, reliability validation

*risk-profile

Generate risk assessment matrix

*risk-profile story-1.2.3

Probability × Impact analysis

*create-fix-request

Generate QA_FIX_REQUEST.md

*create-fix-request story-1.2.3

Creates fix request for @dev with issues

Enhanced Validation

*validate-libraries

Context7 library validation

*validate-libraries story-1.2.3

Validate third-party library usage

*security-check

8-point security scan

*security-check story-1.2.3

Comprehensive security vulnerability scan

*validate-migrations

Database migration validation

*validate-migrations story-1.2.3

Validate schema changes for safety

*evidence-check

Evidence-based QA verification

*evidence-check story-1.2.3

Verify evidence requirements met

*false-positive-check

Critical thinking for bug fixes

*false-positive-check story-1.2.3

Detect false positive bug claims

*console-check

Browser console error detection

*console-check story-1.2.3

Scan browser console for errors

Test Strategy

*test-design

Create comprehensive test scenarios

*test-design story-1.2.3

Design test strategy and scenarios

*trace

Map requirements to tests

*trace story-1.2.3

Given-When-Then traceability mapping

*create-suite

Create test suite for story

*create-suite story-1.2.3

Authority: QA owns test suites

*critique-spec

Review and critique specification

*critique-spec story-1.2.3

Assess spec completeness and clarity

Backlog Management

*backlog-add

*backlog-add story-1.2.3 tech-debt HIGH "Refactor auth"

*backlog-update

*backlog-update item-123 resolved

*backlog-review

*backlog-review

Generate backlog review for sprint

Quality Review Workflow

Standard Review Process

CodeRabbit Self-Healing

Full Self-Healing Loop (Story 6.3.3):
  trigger: review_start
  max_iterations: 3
  timeout_minutes: 30
  severity_filter: [CRITICAL, HIGH]
  
  WHILE iteration < 3:
    1. Run: CodeRabbit --base main
    2. Filter: CRITICAL and HIGH issues
    
    IF no CRITICAL/HIGH issues:
      - Create tech debt for MEDIUM issues
      - Log: "✅ QA passed"
      - BREAK
    
    IF CRITICAL or HIGH issues:
      - Auto-fix each CRITICAL issue
      - Auto-fix each HIGH issue
      - iteration++
      - CONTINUE
  
  IF iteration == 3 AND issues remain:
    - Generate detailed QA gate report
    - Set gate decision: FAIL
    - HALT - require human intervention

Gate Decision Matrix

PASS - Ready to Merge

Criteria:

All acceptance criteria met
No CRITICAL or HIGH CodeRabbit issues
Test coverage adequate
NFRs validated
No blocking security concerns

Action: Approve for merge

CONCERNS - Merge with Notes

Criteria:

Acceptance criteria met
Minor issues or tech debt identified
MEDIUM severity issues
Non-blocking improvements needed

Action: Document concerns, allow merge, create follow-up items

FAIL - Requires Fixing

Criteria:

CRITICAL CodeRabbit issues remain
Acceptance criteria not met
Failing tests
Security vulnerabilities
Major NFR violations

Action: Block merge, create QA_FIX_REQUEST.md, send back to @dev

WAIVED - Documented Risk

Criteria:

Known issues accepted by stakeholders
Technical debt approved for later fix
Risk documented and understood

Action: Document waiver reason, allow merge, create tracking issue

CodeRabbit Integration

Severity Handling

Severity	Action	Focus
CRITICAL	Block story completion	Security vulnerabilities, data integrity risks, critical anti-patterns
HIGH	Report in QA gate, recommend fix	Performance bottlenecks, scalability issues, major anti-patterns
MEDIUM	Document as technical debt	Code maintainability, design patterns, developer experience
LOW	Optional improvements	Style consistency, minor optimizations

Commands

# Pre-review uncommitted
wsl bash -c 'cd ${PROJECT_ROOT} && ~/.local/bin/coderabbit --prompt-only -t uncommitted'

# Story review committed (vs main)
wsl bash -c 'cd ${PROJECT_ROOT} && ~/.local/bin/coderabbit --prompt-only -t committed --base main'

Git Restrictions

@qa is READ-ONLY for git operations

✅ git status, git log, git diff (review only)
✅ git branch -a (list branches)
❌ git push (ONLY @github-devops)
❌ git commit (QA reviews, doesn’t commit)
❌ gh pr create (ONLY @github-devops)

Redirect message: “QA provides advisory review only. For git operations, use appropriate agent (@dev for commits, @github-devops for push)“

Collaboration

Receives work from:

@dev (Dex) - Reviews code and provides feedback

Delegates to:

@dev (Dex) - Sends back via *create-fix-request if FAIL

Collaborates with:

CodeRabbit - Automated code review integration

Usage Examples

@qa
*review story-1.2.3
# Full quality analysis with gate decision

Common Pitfalls

Avoid these common mistakes:

❌ Reviewing before CodeRabbit scan completes
❌ Modifying story sections outside QA Results
❌ Skipping non-functional requirement checks
❌ Not documenting concerns in gate file
❌ Approving without verifying test coverage

@dev (Dex)

Receives feedback from @qa

@sm (River)

May request risk profiling from @qa

CodeRabbit

Automated pre-review integration

Get Started

Core Concepts

Agents

CLI Reference

Workflows

Advanced

​Overview

When to Use @qa

​Agent Profile

​Key Responsibilities

​Available Commands

​Code Review & Analysis

*code-review

*review

*review-build

​Quality Gates

*gate

*nfr-assess

*risk-profile

*create-fix-request

​Enhanced Validation

*validate-libraries

*security-check

*validate-migrations

*evidence-check

*false-positive-check

*console-check

​Test Strategy

*test-design

*trace

*create-suite

*critique-spec

​Backlog Management

*backlog-add

*backlog-update

*backlog-review

​Quality Review Workflow

​Standard Review Process

​CodeRabbit Self-Healing

​Gate Decision Matrix

​CodeRabbit Integration

​Severity Handling

​Commands

​Git Restrictions

​Collaboration

​Receives work from:

​Delegates to:

​Collaborates with:

​Usage Examples

​Common Pitfalls

​Related Agents

@dev (Dex)

@sm (River)

CodeRabbit

Build docs developers (and LLMs) love

Overview

Agent Profile

Key Responsibilities

Available Commands

Code Review & Analysis

Quality Gates

Enhanced Validation

Test Strategy

Backlog Management

Quality Review Workflow

Standard Review Process

CodeRabbit Self-Healing

Gate Decision Matrix

CodeRabbit Integration

Severity Handling

Commands

Git Restrictions

Collaboration

Receives work from:

Delegates to:

Collaborates with:

Usage Examples

Common Pitfalls

Related Agents