Checklists

Overview

AIOX checklists are quality gate validation frameworks used by agents to ensure work meets standards before proceeding. Checklists are markdown files with structured validation criteria, organized by domain and purpose.

string

.aiox-core/development/checklists/ - Development checklists
.aiox-core/product/checklists/ - Product/quality checklists

string

Markdown with checkbox lists and validation instructions

string

Quality gates for tasks, stories, and releases

Checklist Categories

Development Checklists

checklist

Location: .aiox-core/development/checklists/Purpose: Validate agent definition quality before releaseCategories:

Agent metadata completeness
Persona profile validation
Command structure verification
Dependency integrity
Documentation quality
Voice DNA authenticity

Quality Gate: SC_AGT_001 (must pass)

checklist

Purpose: Validate AIOX compatibility with existing (brownfield) projectsValidates:

Existing project structure
Technology stack compatibility
Integration points
Migration path clarity
Risk assessment

checklist

Purpose: Audit agent memory file quality and completenessValidates:

Memory file structure
Entry completeness
Pattern validity
Gotcha accuracy
Decision log integrity

checklist

Purpose: Mandatory self-review for Coder Agent (ADE workflow)Used in: Steps 5.5 & 6.5 of 13-step subtask executionValidates:

Code quality
Requirements alignment
Edge case handling
Error scenarios
Test coverage
Documentation updates

Product Checklists

checklist

Location: .aiox-core/product/checklists/Purpose: Comprehensive architecture validation frameworkSections:

Requirements Alignment
Architecture Fundamentals
Technical Stack & Decisions
Frontend Design & Implementation (if applicable)
Resilience & Operational Readiness
Security & Compliance
Implementation Guidance
Dependency & Integration Management
AI Agent Implementation Suitability
Accessibility Implementation (frontend only)

Total Items: 100+ validation criteriaExecution Modes:

Section-by-section (interactive)
All-at-once (comprehensive)

checklist

Purpose: Validate React/TypeScript component qualityCategories:

Code Quality (TypeScript, props, types)
Styling (Tailwind, variants, dark mode)
Accessibility (WCAG AA, ARIA, keyboard)
Testing (unit, integration, coverage ≥85%)
Documentation (component docs, examples)
Storybook (if enabled)

Standard: Production-ready components

checklist

Purpose: Database schema design validationValidates:

Schema design patterns
Normalization level
Index strategy
RLS (Row-Level Security) policies
Migration integrity
Performance considerations

checklist

Purpose: WCAG 2.2 AA compliance validationValidates:

Semantic HTML
ARIA attributes
Keyboard navigation
Focus management
Color contrast (WCAG & APCA)
Screen reader compatibility

Target: WCAG 2.2 Level AA

checklist

Purpose: Validate changes before commit/mergeCategories:

Code review items
Test coverage
Breaking changes
Documentation updates
Backward compatibility

Checklist Structure

Checklists follow a standard format:

# {Checklist Name}

**Purpose:** {one-line purpose}
**Agent:** {primary-agent}
**Standard:** {quality-standard}

---

## {CATEGORY 1}

- [ ] {validation-criterion-1}
- [ ] {validation-criterion-2}
- [ ] {validation-criterion-3}

---

## {CATEGORY 2}

- [ ] {validation-criterion-1}
  - Sub-criterion details
  - Validation method
- [ ] {validation-criterion-2}

---

**Reviewer:** ________ **Date:** ________
**Quality Gate:** [ ] PASS [ ] FAIL

Checklist Metadata

string

required

One-line description of what the checklist validatesExample: "Validate component before marking complete"

string

Primary agent responsible for checklist executionExample: "Atlas (Design System Builder)", "Quinn (QA Specialist)"

string

Quality standard being enforcedExample: "Production-ready React/TypeScript components", "WCAG 2.2 AA"

Validation Criteria Format

Basic Criteria

format

- [ ] Criterion description

Binary pass/fail validation

Detailed Criteria

format

- [ ] Main criterion
  - Additional detail 1
  - Additional detail 2
  - Validation method: {how-to-verify}

Includes validation guidance

Conditional Criteria

format

- [ ] Criterion (if applicable)
  - Only required when: {condition}
  - Skip if: {skip-condition}

Context-dependent validation

LLM Instructions

format

[[LLM: INSTRUCTION FOR AI AGENT

Before evaluating this section:
1. {guidance-step-1}
2. {guidance-step-2}

Look for specific evidence, not just mentions.]]

Embedded instructions for AI agents

Checklist Execution

Manual Execution

Developers execute checklists manually:

Open checklist file
Work through each criterion
Check boxes as validated
Note any failures or concerns
Complete reviewer/date fields
Mark overall PASS/FAIL

Agent Execution

Agents execute checklists programmatically:

# Execute checklist via command
*execute-checklist component-quality-checklist

# Agent workflow:
# 1. Load checklist file
# 2. Read [[LLM: ...]] instructions
# 3. Validate each criterion
# 4. Generate validation report
# 5. Return PASS/FAIL with details

Integration with Tasks

Checklists are invoked by tasks:

# From dev-develop-story.md
completion: |
  All Tasks marked [x] →
  Validations pass →
  Run task execute-checklist for story-dod-checklist →
  Set story status: 'Ready for Review' →
  HALT

Quality Gates

concept

A quality gate is a checkpoint that must pass before proceedingImplementation:

Checklist defines criteria
Agent validates all items
Gate passes only if all criteria met
On failure: Block progression, provide report

Quality Gate Standards

gate

Agent Quality Gate - All agents must pass before releaseValidates:

300+ lines of content
Voice DNA authenticity
Output examples included
Command loader present
Task files exist
Maturity level ≥ Nivel 2

gate

Story Definition of Done - All stories must pass before reviewValidates:

All tasks completed
Tests passing
Documentation updated
File list complete
No regression failures

Advanced Checklist Features

Project Type Detection

[[LLM: PROJECT TYPE DETECTION

First, determine the project type:
- Does architecture include frontend/UI?
- Is there a frontend-architecture.md?

If backend-only:
- Skip sections marked [[FRONTEND ONLY]]
- Focus on API design, service architecture
- Note in final report]]

## Frontend Architecture [[FRONTEND ONLY]]

[[LLM: Skip this entire section for backend-only projects]]

- [ ] UI framework selected
- [ ] State management defined

Evidence-Based Validation

[[LLM: VALIDATION APPROACH

For each section:
1. Deep Analysis - Don't just check boxes
2. Evidence-Based - Cite specific sections
3. Critical Thinking - Question assumptions
4. Risk Assessment - Consider what could go wrong]]

Execution Modes

[[LLM: EXECUTION MODE

Ask user preference:
- Section by section (interactive) - Review each, get confirmation
- All at once (comprehensive) - Complete full analysis, present report]]

Final Report Generation

[[LLM: FINAL VALIDATION REPORT

Generate comprehensive report:

1. Executive Summary
   - Overall readiness (High/Medium/Low)
   - Critical risks
   - Key strengths

2. Section Analysis
   - Pass rate per section
   - Most concerning failures
   - Sections needing attention

3. Risk Assessment
   - Top 5 risks by severity
   - Mitigation recommendations

4. Recommendations
   - Must-fix before development
   - Should-fix for quality
   - Nice-to-have improvements]]

Checklist Best Practices

Checklist Design

Keep criteria clear and actionable
Provide validation methods
Group related items into sections
Include LLM instructions for AI agents
Support conditional validation
Define clear pass/fail criteria

Validation Criteria

Make criteria measurable
Avoid subjective judgments
Provide examples where helpful
Include specific thresholds (e.g., “≥85% coverage”)
Reference standards (e.g., “WCAG 2.2 AA”)

Agent Integration

Use [[LLM: …]] blocks for agent guidance
Provide context and reasoning
Include validation examples
Define report format expectations
Support both manual and automated execution

Maintenance

Review checklists after each usage
Update based on team feedback
Add new criteria as standards evolve
Remove obsolete criteria
Version checklists when making changes

Creating Custom Checklists

Choose location:

# Development checklist
touch .aiox-core/development/checklists/my-checklist.md

# Product checklist
touch .aiox-core/product/checklists/my-checklist.md

Define structure:

# My Quality Checklist

**Purpose:** Validate {what}
**Agent:** {agent-name}
**Standard:** {quality-level}

---

## Category 1

- [ ] Criterion 1
- [ ] Criterion 2

---

**Quality Gate:** [ ] PASS [ ] FAIL

Register with agent:

dependencies:
  checklists:
    - my-checklist.md

Integrate with task:

## Quality Gate

Execute checklist: `my-checklist.md`

Must pass before proceeding to next phase.

Reference: .aiox-core/development/checklists/ and .aiox-core/product/checklists/

Configuration

Agent System

Templates & Tasks

Overview

Checklist Categories

Development Checklists

Product Checklists

Checklist Structure

Checklist Metadata

Validation Criteria Format

Basic Criteria

Detailed Criteria

Conditional Criteria

LLM Instructions

Checklist Execution

Manual Execution

Agent Execution

Integration with Tasks

Quality Gates

Quality Gate Standards

Advanced Checklist Features

Project Type Detection

Evidence-Based Validation

Execution Modes

Final Report Generation

Checklist Best Practices

Creating Custom Checklists

​Overview

​Checklist Categories

​Development Checklists

​Product Checklists

​Checklist Structure

​Checklist Metadata

​Validation Criteria Format

​Basic Criteria

​Detailed Criteria

​Conditional Criteria

​LLM Instructions

​Checklist Execution

​Manual Execution

​Agent Execution

​Integration with Tasks

​Quality Gates

​Quality Gate Standards

​Advanced Checklist Features

​Project Type Detection

​Evidence-Based Validation

​Execution Modes

​Final Report Generation

​Checklist Best Practices

​Creating Custom Checklists

​Related Documentation

Overview

Checklist Categories

Development Checklists

Product Checklists

Checklist Structure

Checklist Metadata

Validation Criteria Format

Basic Criteria

Detailed Criteria

Conditional Criteria

LLM Instructions

Checklist Execution

Manual Execution

Agent Execution

Integration with Tasks

Quality Gates

Quality Gate Standards

Advanced Checklist Features

Project Type Detection

Evidence-Based Validation

Execution Modes

Final Report Generation

Checklist Best Practices

Creating Custom Checklists

Related Documentation