Skip to content

COMMANDS_GUIDE.md

Latest commit

 

History

History
1037 lines (681 loc) · 20.5 KB

COMMANDS_GUIDE.md

File metadata and controls

1037 lines (681 loc) · 20.5 KB

AIWF (AI Workflow Framework) Commands Guide

This guide explains the available AIWF commands and how to use them effectively in your projects.

Multi-language Support: AIWF supports both Korean and English. Commands and documentation are provided in the language selected during installation.

Command Overview

AIWF commands follow the format /aiwf_<command_name> [arguments].

Note:

  • The /project:aiwf:<command_name> format is still supported for backward compatibility.
  • Korean commands may have a _kr suffix (e.g., /aiwf_initialize_kr)

Language Management System

Language Support Features

AIWF includes a comprehensive language management system with:

  • Intelligent Language Detection: Automatically detects system language preferences
  • Persistent Language Storage: Saves your language preference for future installations
  • Interactive Language Switching: Switch between languages with simple commands
  • Command-Line Language Tools: Dedicated CLI tools for language management

Language Commands

🌐 /aiwf_language_manager

Purpose: Comprehensive language management interface

Usage:

/aiwf_language_manager

Behavior:

  • Shows current language settings and preferences
  • Provides options to change language interactively
  • Manages language detection and storage
  • Validates language settings across the system

🔍 /aiwf_language_status

Purpose: Check current language configuration and status

Usage:

/aiwf_language_status

Behavior:

  • Displays current active language
  • Shows detected system language preferences
  • Reports language file availability and consistency
  • Provides language-specific installation status

🔄 /aiwf_switch_language

Purpose: Switch between supported languages

Usage:

/aiwf_switch_language
# Interactive language selection

/aiwf_switch_language ko
# Switch to Korean

/aiwf_switch_language en
# Switch to English

Behavior:

  • Updates language preferences
  • Reinstalls commands in selected language
  • Maintains project settings and data
  • Provides confirmation of successful language switch

Command Versions

  • Default Commands: English or standardized multilingual
  • Korean Commands: Complete Korean versions with full feature parity
  • Language Management: Available in both languages with identical functionality

Setup & Context Commands

🚀 /aiwf_initialize

Purpose: Initialize AIWF in a new or existing project

Usage:

/aiwf_initialize

Behavior:

  1. Scans and analyzes the project
  2. Requests project type confirmation
  3. Checks for existing AIWF documents
  4. Guides document creation (import existing or create new)
  5. Creates first milestone
  6. Generates project manifest

When to Use: When first setting up AIWF in a project

🧠 /aiwf_prime

Purpose: Load project context at the start of a coding session

Usage:

/aiwf_prime

Behavior:

  • Reads project manifest
  • Loads current milestone and sprint information
  • Identifies active tasks
  • Provides quick status overview

When to Use: At the start of a coding session to quickly understand the overall situation

Planning Commands

📅 /aiwf_create_sprints_from_milestone

Purpose: Break down milestones into manageable sprints

Usage:

/aiwf_create_sprints_from_milestone 001_MVP_FOUNDATION

Behavior:

  1. Analyzes milestone requirements
  2. Groups related requirements into ~1 week sprints
  3. Creates sprint folders and META files
  4. Updates sprint information in manifest

When to Use: After creating a new milestone

📋 /aiwf_create_sprint_tasks

Purpose: Create detailed task breakdown for a sprint

Usage:

/aiwf_create_sprint_tasks S01
# or for specific sprint:
/aiwf_create_sprint_tasks S02_001_MVP_FOUNDATION

Behavior:

  1. Analyzes sprint requirements
  2. Breaks them into concrete, actionable tasks
  3. Creates task files with clear goals
  4. Handles inter-task dependencies

When to Use: At the start of each sprint

✏️ /aiwf_create_general_task

Purpose: Create standalone tasks not tied to sprints

Usage:

/aiwf_create_general_task
# Then describe your task when prompted

Example Tasks:

  • "Fix memory leak in physics engine"
  • "Update documentation for API changes"
  • "Refactor database connection pooling"

When to Use: For maintenance, bug fixes, or work outside sprint scope

Development Commands

💻 /aiwf_do_task

Purpose: Execute a specific task

Usage:

/aiwf_do_task
# Lists available tasks and prompts for selection

# Or specify task directly:
/aiwf_do_task T001_S01_setup_tauri

Behavior:

  1. Read task requirements
  2. Implement solution
  3. Run tests if applicable
  4. Update task status
  5. Create necessary files/changes

When to Use: When ready to work on a specific task

📝 /aiwf_commit

Purpose: Create well-structured git commits and link with GitHub issues

Usage:

/aiwf_commit
# Reviews changes and creates commit

# Or for specific task:
/aiwf_commit T001_S01_setup_tauri

# With review:
/aiwf_commit --review

Behavior:

  1. Analyze changes
  2. Group related changes
  3. Generate meaningful commit messages
  4. Link commits to tasks/requirements
  5. Integrate with GitHub issues (fixes #123, relates to #456)
  6. Optionally run code review first

When to Use: When you want to save completed work

🧪 /aiwf_test

Purpose: Run tests and fix common issues

Usage:

/aiwf_test
# Run all tests

/aiwf_test unit
# Run specific test suite

Behavior:

  1. Identify test commands from package.json
  2. Run appropriate tests
  3. Fix common issues (missing dependencies, config)
  4. Report results clearly

When to Use: Before committing or when tests fail

Code Review Commands

🔍 /aiwf_code_review

Purpose: Review code against specifications

Usage:

/aiwf_code_review
# Review uncommitted changes

/aiwf_code_review src/app/components/GameCanvas.tsx
# Review specific file

Behavior:

  1. Check code against requirements
  2. Verify patterns and conventions
  3. Identify bugs and issues
  4. Suggest improvements
  5. Ensure spec compliance

When to Use: Before committing significant changes

📊 /aiwf_project_review

Purpose: Overall project health check

Usage:

/aiwf_project_review

Behavior:

  1. Review overall architecture
  2. Identify technical debt
  3. Analyze progress vs schedule
  4. Identify risks and blockers
  5. Suggest improvements

When to Use: Weekly or at sprint boundaries

🧪 /aiwf_testing_review

Purpose: Analyze test coverage and quality

Usage:

/aiwf_testing_review

Behavior:

  1. Review test coverage
  2. Identify missing test cases
  3. Check test quality
  4. Suggest improvements

When to Use: After implementing features

💬 /aiwf_discuss_review

Purpose: Technical discussion about review findings

Usage:

/aiwf_discuss_review
# After running other review commands

Behavior:

  • Provide detailed explanations
  • Discuss pros and cons
  • Suggest solutions
  • Answer questions

When to Use: To better understand review feedback

Automation Commands

🚀 /aiwf_yolo

Purpose: Autonomous task execution

Usage:

/aiwf_yolo
# Execute all open tasks sequentially

/aiwf_yolo S02
# Execute specific sprint sequentially

/aiwf_yolo sprint-all
# Execute all sprints sequentially

/aiwf_yolo milestone-all
# Execute all milestones sequentially

/aiwf_yolo S02 worktree
# Execute in Git worktree mode

Behavior:

  1. Identify open tasks
  2. Execute in order
  3. Handle dependencies
  4. Commit completed tasks
  5. Update progress
  6. Update GitHub issue status

Safety Features:

  • Won't modify schema without confirmation
  • Skips dangerous operations
  • Maintains code quality
  • Creates logical commits
  • Stops on test failures

When to Use: When you want autonomous progress

CLI Commands (New in v0.3.12)

🚀 aiwf sprint independent

Purpose: Create independent sprints without milestones for rapid YOLO execution

Usage:

# Extract from README TODOs
aiwf sprint independent --from-readme

# Create from GitHub issue
aiwf sprint independent --from-issue 123

# Interactive creation with engineering level
aiwf sprint independent "Quick Feature" --minimal
aiwf sprint independent "API Development" --balanced
aiwf sprint independent "Complex System" --complete

Features:

  • No milestone dependency required
  • Automatic task extraction from README/Issues
  • Configurable engineering levels
  • YOLO-optimized sprint structure

💾 aiwf checkpoint

Purpose: Manage YOLO execution checkpoints for recovery and tracking

Usage:

# List checkpoints
aiwf checkpoint list
aiwf checkpoint list --limit 20

# Check current session status
aiwf checkpoint status

# Restore from checkpoint
aiwf checkpoint restore cp_1234567890

# Create manual checkpoint
aiwf checkpoint create "Before major refactor"

# Clean old checkpoints
aiwf checkpoint clean --keep 10

Features:

  • Automatic checkpoint creation during YOLO
  • Git state tracking
  • Session metrics and performance data
  • Safe recovery from interruptions

🛠️ aiwf yolo-config

Purpose: Configure YOLO behavior and overengineering prevention

Usage:

# Initialize default config
aiwf yolo-config init

# Interactive configuration wizard
aiwf yolo-config wizard

# Show current configuration
aiwf yolo-config show

Configuration Options:

  • Engineering levels (minimal/balanced/complete)
  • Focus rules (requirement_first, simple_solution, etc.)
  • Execution modes (fast/smart/safe)
  • Overengineering prevention limits
  • Checkpoint intervals

📊 aiwf-sprint (Dedicated CLI)

Purpose: Comprehensive sprint management tool

Usage:

# Create independent sprint
aiwf-sprint independent --from-readme --minimal

# List all sprints
aiwf-sprint list
aiwf-sprint list --status active

# Check sprint status
aiwf-sprint status S01

# Get help
aiwf-sprint help

🔄 aiwf-checkpoint (Dedicated CLI)

Purpose: Advanced checkpoint management

Usage:

# Generate progress report
aiwf-checkpoint report

# List with details
aiwf-checkpoint list

# Clean with dry-run
aiwf-checkpoint clean --keep 5 --dry-run

# Get help
aiwf-checkpoint help

Additional/Advanced Commands

📌 /aiwf_pr_create

Purpose: Create Pull Request with templates to organize changes

Usage:

/aiwf_pr_create
# Create PR interactively

/aiwf_pr_create "Add authentication to API"
# Create PR with title

Behavior:

  1. Analyze current branch changes
  2. Identify related issues and tasks
  3. Generate PR title and description
  4. Include test checklist
  5. Create PR on GitHub

🗂️ /aiwf_issue_create

Purpose: Create GitHub Issue for bug reports and feature requests

Usage:

/aiwf_issue_create
# Create issue interactively

/aiwf_issue_create "Bug: login fails on Safari"
# Create issue with title

Behavior:

  1. Get issue title and description
  2. Auto-assign appropriate labels and milestones
  3. Apply bug/feature request templates
  4. Create issue on GitHub
  5. Convert to task if needed

🛠️ /aiwf_create_milestone_plan

Purpose: Plan new milestone through interactive process and auto-generate .aiwf/02_REQUIREMENTS/ structure

Usage:

/aiwf_create_milestone_plan

Behavior:

  1. Define milestone goals and scope
  2. Create requirements document structure
  3. Generate PRD and technical spec templates
  4. Set up milestone directory structure
  5. Update project manifest

📝 /aiwf_create_prd

Purpose: Create Product Requirements Document (PRD) with detailed feature specifications

Usage:

/aiwf_create_prd
# Create PRD interactively

/aiwf_create_prd "User Authentication System"
# Create PRD for specific feature

Behavior:

  1. Define feature purpose and scope
  2. Write user stories and scenarios
  3. Specify technical requirements
  4. Identify constraints and dependencies
  5. Create PRD document structure

When to Use: When detailed specifications are needed before implementing new features

📈 /aiwf_mermaid

Purpose: Generate Mermaid diagrams by analyzing codebase

Usage:

/aiwf_mermaid
# Generate overall architecture diagram

/aiwf_mermaid flowchart
# Generate flowchart

/aiwf_mermaid sequence
# Generate sequence diagram

Behavior:

  1. Analyze codebase structure
  2. Identify component and module relationships
  3. Select appropriate diagram type
  4. Generate diagram in Mermaid syntax
  5. Output in document-ready format

♾️ /aiwf_infinite

Purpose: Run advanced iteration loop to generate outputs repeatedly according to specifications

Usage:

/aiwf_infinite
# Infinite iteration mode

/aiwf_infinite 5
# 5 iteration mode

Behavior:

  1. Define iteration generation rules
  2. Set initial conditions
  3. Execute specified number of iterations
  4. Validate results each iteration
  5. Integrate final outputs

🤖 /aiwf_tm-run-all-subtask

Purpose: Execute all Task Master subtasks at once to automate progress

Usage:

/aiwf_tm-run-all-subtask

Behavior:

  1. Identify active subtasks
  2. Determine dependency order
  3. Judge parallel execution feasibility
  4. Execute subtasks sequentially
  5. Update overall progress

🧠 /aiwf_ultrathink_general

Purpose: Ultra thinking session for deep analysis of broad problems

Usage:

/aiwf_ultrathink_general "Complex business logic design"

🧠 /aiwf_ultrathink_code_basic

Purpose: Ultra thinking session for deep analysis of code-based problems at basic level

Usage:

/aiwf_ultrathink_code_basic "Performance optimization approaches"

🧠 /aiwf_ultrathink_code_advanced

Purpose: Ultra thinking session for deep analysis of complex code and architecture problems at advanced level

Usage:

/aiwf_ultrathink_code_advanced "Microservices architecture design"

⚙️ /aiwf_update_docs

Purpose: Update and synchronize project documentation

Usage:

/aiwf_update_docs

Behavior:

  1. Scan project documentation structure
  2. Identify outdated or missing documentation
  3. Update documentation to reflect current project state
  4. Synchronize documentation across different formats
  5. Validate documentation consistency and completeness

⚙️ /aiwf_prime_context

Purpose: Quickly load and prime project context

Usage:

/aiwf_prime_context

Behavior:

  1. Scan project file list
  2. Identify key documents
  3. Summarize current state
  4. Load context information
  5. Set task-ready state

AI Persona Commands

🎭 /project:aiwf:ai_persona:switch

Purpose: Switch to a specific AI persona for focused development tasks

Usage:

/project:aiwf:ai_persona:switch architect
# Switch to architect persona

/project:aiwf:ai_persona:switch debugger
# Switch to debugger persona

Available Personas:

  • architect - System design and architecture
  • debugger - Bug detection and troubleshooting
  • reviewer - Code quality and standards
  • documenter - Documentation and guides
  • optimizer - Performance optimization
  • developer - General development (default)

🎭 /project:aiwf:ai_persona:architect

Purpose: Shortcut to switch to architect persona

Usage:

/project:aiwf:ai_persona:architect

When to Use: When designing system architecture, planning modules, creating technical designs

🎭 /project:aiwf:ai_persona:debugger

Purpose: Shortcut to switch to debugger persona

Usage:

/project:aiwf:ai_persona:debugger

When to Use: When fixing bugs, tracing errors, investigating issues

🎭 /project:aiwf:ai_persona:reviewer

Purpose: Shortcut to switch to reviewer persona

Usage:

/project:aiwf:ai_persona:reviewer

When to Use: When reviewing code, checking security, ensuring standards

🎭 /project:aiwf:ai_persona:documenter

Purpose: Shortcut to switch to documenter persona

Usage:

/project:aiwf:ai_persona:documenter

When to Use: When writing documentation, creating guides, explaining features

🎭 /project:aiwf:ai_persona:optimizer

Purpose: Shortcut to switch to optimizer persona

Usage:

/project:aiwf:ai_persona:optimizer

When to Use: When optimizing performance, reducing resource usage, improving efficiency

🔄 /project:aiwf:ai_persona:auto

Purpose: Enable or disable automatic persona detection

Usage:

/project:aiwf:ai_persona:auto on
# Enable auto-detection

/project:aiwf:ai_persona:auto off
# Disable auto-detection

Behavior: When enabled, Claude automatically switches personas based on task context

📊 /project:aiwf:ai_persona:status

Purpose: Display current persona and session metrics

Usage:

/project:aiwf:ai_persona:status

Output: Shows current persona, behaviors, session duration, interactions, and token usage

📋 /project:aiwf:ai_persona:list

Purpose: List all available AI personas

Usage:

/project:aiwf:ai_persona:list

Output: Shows all personas with descriptions and focus areas

🔍 /project:aiwf:ai_persona:detect

Purpose: Detect optimal persona for a specific task

Usage:

/project:aiwf:ai_persona:detect "fix the authentication bug"
# Detects debugger persona

/project:aiwf:ai_persona:detect "design the API structure"
# Detects architect persona

📈 /project:aiwf:ai_persona:report

Purpose: Generate performance report for persona usage

Usage:

/project:aiwf:ai_persona:report
# Generate report for all time

/project:aiwf:ai_persona:report week
# Generate report for past week

Output: Performance metrics, persona usage statistics, recommendations

📊 /project:aiwf:ai_persona:stats

Purpose: View detailed statistics for a specific persona

Usage:

/project:aiwf:ai_persona:stats debugger
# View debugger persona stats

/project:aiwf:ai_persona:stats
# View current persona stats

🔄 /project:aiwf:ai_persona:reset

Purpose: Reset persona system to default (developer)

Usage:

/project:aiwf:ai_persona:reset

Best Practices

Daily Workflow

# Start of day
/aiwf_prime

# Enable AI persona auto-detection
/project:aiwf:ai_persona:auto on

# Work on tasks
/aiwf_do_task
/aiwf_test
/aiwf_commit

# End of day
/aiwf_project_review
/project:aiwf:ai_persona:report

Sprint Workflow

# Sprint planning
/aiwf_create_sprint_tasks S02

# Sprint execution
/aiwf_do_task T001_S02_first_task
/aiwf_do_task T002_S02_second_task
/aiwf_commit --review

# Sprint review
/aiwf_project_review

Quick Fixes

# Bug fix workflow
/aiwf_create_general_task
# Description: "Fix memory leak in /src/foo.bar"
/aiwf_do_task T003
/aiwf_test
/aiwf_commit T003

Tips & Tricks

  1. Use YOLO for routine work: Useful for simple feature implementation
  2. Always run prime first: Ensures commands have proper context
  3. Review before major commits: Catch issues early
  4. Create bugs as general tasks: Keep them trackable
  5. Use task-specific commits: Better traceability
  6. Enable AI persona auto-detection: Let Claude automatically switch to the right mode
  7. Use architect persona for planning: Better system design decisions
  8. Switch to debugger for bug fixing: More systematic error analysis
  9. Use optimizer persona sparingly: Focus on actual bottlenecks
  10. Check persona reports weekly: Understand your development patterns

Command Safety Features

AIWF commands include the following safety features:

  • Won't delete important files
  • Requests confirmation before schema changes
  • Validates changes against specifications
  • Maintains code quality standards
  • Creates incremental commits

Getting Help

If you need help with commands:

  1. Run commands without arguments to see usage info
  2. Check this guide
  3. Look at task examples in .aiwf/
  4. Review command sources in .claude/commands/

Document Updates

📝 Important: This document is automatically updated when new features are added or existing commands are modified. Always refer to the latest version for accurate command information.

Language Versions: Both English and Korean versions of this guide are maintained and updated simultaneously to ensure consistency across language installations.

Last Updated: 2025-07-09 - AI Persona System Integration and Documentation