### Execute Project Setup Steps

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif/SKILL.md

This bash script outlines the steps to execute the project setup, including directory creation, configuration updates, and skill installation.

```bash
mkdir -p .ai-factory
node ~/{{skills_dir}}/aif/references/update-config.mjs --template ~/{{skills_dir}}/aif/references/config-template.yaml --target .ai-factory/config.yaml --payload .ai-factory/config.update.json
rm .ai-factory/config.update.json
# Ensure .ai-factory/rules/ directory exists
# Write .ai-factory/rules/base.md
# ... other steps
```

--------------------------------

### Docker Setup Summary

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-dockerize/references/SUMMARY-FORMAT.md

This is a template for summarizing a Docker setup. It lists created files, provides quick start commands, and details services.

```markdown
## Docker Setup Complete

### Files Created/Updated
- Dockerfile (multi-stage: development + production)
- compose.yml (app + postgres + redis, COMPOSE_PROJECT_NAME from .env)
- compose.override.yml (dev: hot reload, debug ports, mailpit)
- compose.production.yml (hardened: read-only, non-root, resource limits, no infra ports)
- .dockerignore (38 exclusion rules)
- .env.example (with COMPOSE_PROJECT_NAME, DB credentials, app config)
- docker/angie/ (reverse proxy config, if needed)
- deploy/scripts/ (deploy, update, logs, health-check, rollback, backup)

### Quick Start
  # Development
  docker compose up

  # Development with email testing
  docker compose --profile dev up

  # Production (locally)
  docker compose -f compose.yml -f compose.production.yml up -d

  # Build production image
  docker build --target production -t myapp:latest .

### Services
| Service | Port (dev) | Port (prod) | Image |
|---------|------------|-------------|-------|
| app | 3000, 9229 | — | built locally |
| postgres | 5432 | — | postgres:17-alpine |
| redis | 6379 | — | redis:7-alpine |
| mailpit | 8025, 1025 | — | axllent/mailpit |
```

--------------------------------

### Installation Hint Format

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-build-automation/references/SUMMARY-FORMAT.md

A general format for providing installation instructions for build tools if they are not already installed. This ensures users can set up the necessary environment.

```markdown
### Installation
  [install instructions for task/just/mage if not already installed]
```

--------------------------------

### Go Build Command Example

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-ci/SKILL.md

Example of a build command for Go projects.

```bash
go build ./...
```

--------------------------------

### Java Build Command Examples

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-ci/SKILL.md

Examples of build commands for Java projects using Maven or Gradle.

```bash
mvn package -DskipTests -B
/ ./gradlew assemble
```

--------------------------------

### Rust Build Command Example

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-ci/SKILL.md

Example of a release build command for Rust projects.

```bash
cargo build --release
```

--------------------------------

### Example Extension Directory Structure

Source: https://github.com/lee-to/ai-factory/blob/2.x/docs/extensions.md

This illustrates the typical directory layout for an installed AI Factory extension.

```tree
your-project/
├── .ai-factory/
│   ├── extensions/
│   │   └── aif-ext-example/        # Installed extension
│   │       ├── extension.json
│   │       ├── commands/
│   │       ├── injections/
│   │       ├── skills/
│   │       └── mcp/
│   └── ...
├── .ai-factory.json                 # extensions[] array tracks installed extensions
└── .mcp.json                        # MCP servers merged here (for Claude Code)
```

--------------------------------

### Just Installation Command

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-build-automation/references/SUMMARY-FORMAT.md

Provides the command to install the Just build tool using Cargo or Homebrew.

```bash
cargo install just
```

```bash
brew install just
```

--------------------------------

### Standard MCP Runtime Wrapper Example

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif/SKILL.md

Example of wrapping the filesystem server for standard MCP runtimes, using the 'mcpServers' key.

```json
{
  "mcpServers": {
    "filesystem": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-filesystem", "."]
    }
  }
}
```

--------------------------------

### Make Installation Command

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-build-automation/references/SUMMARY-FORMAT.md

Provides the command to install GNU Make, typically pre-installed on most systems.

```bash
brew install make
```

--------------------------------

### Example .env.example File

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-dockerize/references/FILE-ORGANIZATION.md

This example `.env.example` file demonstrates how to define project, database, application, and production-specific environment variables. Production variables are commented out and can be uncommented for deployment.

```env
# === Project ===
COMPOSE_PROJECT_NAME=myapp

# === Database ===
DB_NAME=mydb
DB_USER=app
DB_PASSWORD=changeme
POSTGRES_VERSION=17

# === Application ===
LOG_LEVEL=debug                          # prod: warn

# (add project-specific vars detected in Step 2.8)

# === Production (uncomment for deploy) ===
# DOCKER_REGISTRY=ghcr.io
# DOCKER_IMAGE=myapp
# VERSION=latest
# ALLOWED_ORIGINS=https://myapp.com
# TRUSTED_PROXIES=172.16.0.0/12
```

--------------------------------

### OpenCode Runtime Wrapper Example

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif/SKILL.md

Example of wrapping the filesystem server for OpenCode runtime. Note the 'type: "local"' and command as an array.

```json
{
  "mcp": {
    "filesystem": {
      "type": "local",
      "command": ["npx", "-y", "@modelcontextprotocol/server-filesystem", "."]
    }
  }
}
```

--------------------------------

### Example Extension Manifest with Skill Replacement

Source: https://github.com/lee-to/ai-factory/blob/2.x/docs/extensions.md

This example shows a typical extension manifest structure including custom skills and a skill replacement for `aif-commit`. The extension's `my-commit/SKILL.md` will be installed as `aif-commit/SKILL.md`.

```json
{
  "name": "aif-ext-custom-commit",
  "version": "1.0.0",
  "skills": ["skills/my-commit"],
  "replaces": {
    "skills/my-commit": "aif-commit"
  }
}
```

--------------------------------

### Task Installation Command

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-build-automation/references/SUMMARY-FORMAT.md

Provides the command to install the Task build tool using Go modules or Homebrew.

```bash
go install github.com/go-task/task/v3/cmd/task@latest
```

```bash
brew install go-task
```

--------------------------------

### Resuming Implementation Example

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-implement/references/IMPLEMENTATION-GUIDE.md

Example dialogue showing the agent resuming an implementation session, detecting completed tasks, and continuing from the next pending task.

```text
User: /aif-implement

Agent: Resuming implementation...

Found 3 completed tasks, 5 pending.
Continuing from Task #4: Implement JWT generation

[Executes task #4]
```

--------------------------------

### GitHub Copilot Runtime Wrapper Example

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif/SKILL.md

Example of wrapping the filesystem server for GitHub Copilot runtime. Note the 'type: "stdio"' key.

```json
{
  "servers": {
    "filesystem": {
      "type": "stdio",
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-filesystem", "."]
    }
  }
}
```

--------------------------------

### Example .env Files

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-dockerize/references/BEST-PRACTICES.md

Provides examples of `.env` files for managing environment variables. `.env` contains real secrets and is gitignored, while `.env.example` contains placeholders and is committed.

```env
# .env (gitignored, never committed) — real values
COMPOSE_PROJECT_NAME=myapp
DB_PASSWORD=real-secret-value
OPENAI_API_KEY=sk-...
```

```env
# .env.example (committed, with placeholders)
COMPOSE_PROJECT_NAME=myapp
DB_PASSWORD=changeme
OPENAI_API_KEY=
```

--------------------------------

### Install and Initialize AI Factory Project

Source: https://github.com/lee-to/ai-factory/blob/2.x/docs/getting-started.md

Install AI Factory globally, navigate to your project, and initialize it. You can use the interactive wizard or specify agents and MCPs non-interactively.

```bash
# 1. Install AI Factory
npm install -g ai-factory

# 2. Go to your project
cd my-project

# 3. Initialize — pick agents, install skills, configure MCP
ai-factory init
# Or non-interactively: ai-factory init --agents claude --mcp github,playwright

# 4. Open your AI agent (Claude Code, Cursor, etc.) and run:
/aif
# Codex CLI and Codex app use: $aif

# 5. Optional discovery before planning
/aif-explore Add user authentication with OAuth

# 6. Start building
/aif-plan Add user authentication with OAuth
```

--------------------------------

### Node.js Build Command Example

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-ci/SKILL.md

Example of a build command for Node.js projects that have a 'build' script defined in their package.json.

```bash
npm run build
/ pnpm build
```

--------------------------------

### AI Factory Project Setup and Modes

Source: https://github.com/lee-to/ai-factory/blob/2.x/AGENTS.md

The `/aif` command initiates project setup with three distinct modes based on project state and user input: existing projects, new projects with descriptions, and empty projects.

```bash
/aif (3 scenarios)
    ↓
Check: has arguments? has project files?
    ↓
┌─────────────────────────────────────────────────────────────┐
│ Mode 1: Existing project (no args + has package.json, etc.) │
│   → Analyze project → Generate DESCRIPTION.md → Setup       │
├─────────────────────────────────────────────────────────────┤
│ Mode 2: New project with description                        │
│   → Interactive stack selection → DESCRIPTION.md → Setup    │
├─────────────────────────────────────────────────────────────┤
│ Mode 3: Empty project (no args + no config files)           │
│   → Ask "What are you building?" → Stack selection → Setup  │
└─────────────────────────────────────────────────────────────┘
```

--------------------------------

### TaskCreate Example

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-plan/references/TASK-FORMAT.md

Example of a TaskCreate entry, specifying the subject, a detailed description including logging requirements, and the active form.

```text
TaskCreate:
  subject: "Implement user login endpoint"
  description: |
    Create POST /api/auth/login endpoint that:
    - Accepts email and password
    - Validates credentials against database
    - Returns JWT token on success
    - Returns 401 on invalid credentials

    LOGGING REQUIREMENTS:
    - Log function entry with request context
    - Log validation result (pass/fail with reasons)
    - Log external service calls and responses
    - Log any errors with full context
    - Use format: [ServiceName.method] message {data}
    - Use log levels (DEBUG/INFO/WARN/ERROR)

    Files: src/api/auth/login.ts, src/services/auth.ts
  activeForm: "Implementing login endpoint"

```

--------------------------------

### AI Factory Configuration Example

Source: https://github.com/lee-to/ai-factory/blob/2.x/docs/configuration.md

This is a comprehensive example of the `.ai-factory/config.yaml` file, illustrating various settings for language, paths, workflow, git integration, and rules. Defaults are used when sections or keys are not specified.

```yaml
# AI Factory Configuration
# All sections are optional — defaults are used when not specified.

# Language Settings
language:
  # Language for AI-agent communication (prompts, questions, explanations)
  # Options: en, ru, de, fr, es, zh, ja, ko, pt, it
  ui: en

  # Language for generated artifacts (plans, specs, documentation)
  artifacts: en

  # How to handle technical terms: keep | translate | mixed
  technical_terms: keep

# Path Configuration (all relative to project root)
paths:
  description: .ai-factory/DESCRIPTION.md
  architecture: .ai-factory/ARCHITECTURE.md
  docs: docs/
  roadmap: .ai-factory/ROADMAP.md
  research: .ai-factory/RESEARCH.md
  rules_file: .ai-factory/RULES.md
  plan: .ai-factory/PLAN.md
  plans: .ai-factory/plans/
  fix_plan: .ai-factory/FIX_PLAN.md
  security: .ai-factory/SECURITY.md
  references: .ai-factory/references/
  patches: .ai-factory/patches/
  evolutions: .ai-factory/evolutions/
  evolution: .ai-factory/evolution/
  specs: .ai-factory/specs/
  rules: .ai-factory/rules/
  qa: .ai-factory/qa/
  archive: .ai-factory/archive/

# Workflow Settings
workflow:
  auto_create_dirs: true           # Create .ai-factory/ directories when missing
  plan_id_format: slug             # slug | sequential (timestamp, uuid — reserved, fall back to slug)
  analyze_updates_architecture: true
  architecture_updates_roadmap: true
  verify_mode: normal              # strict | normal | lenient

# Git Settings
git:
  enabled: true                    # Set false for non-git repositories
  base_branch: main                # Diff / review / merge target when git is enabled
  create_branches: true            # Full plans may create branches when enabled
  branch_prefix: feature/          # Prefix for auto-created plan branches
  skip_push_after_commit: false    # If true, /aif-commit skips push prompt after commit

# Rules Configuration
rules:
  base: .ai-factory/rules/base.md  # Base rules file
  # api: .ai-factory/rules/api.md
  # frontend: .ai-factory/rules/frontend.md
  # backend: .ai-factory/rules/backend.md
  # database: .ai-factory/rules/database.md

```

--------------------------------

### Display Implementation Progress

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-implement/SKILL.md

Shows an example of how implementation progress is displayed to the user.

```markdown
## Implementation Progress

✅ Completed: 3/8 tasks
🔄 In Progress: Task #4 - Implement search service
⏳ Pending: 4 tasks

Current task: #4 - Implement search service
```

--------------------------------

### Docker Compose Volume Mount Example

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-dockerize/references/FILE-ORGANIZATION.md

Example of how to configure volume mounts in `docker-compose.yml` to reference configuration files from the `docker/` directory, such as for an Angie reverse proxy.

```yaml
# Example: Angie config mount
volumes:
  - ./docker/angie/angie.conf:/etc/angie/angie.conf:ro
  - ./docker/angie/conf.d:/etc/angie/conf.d:ro
```

--------------------------------

### Example Invocations for Learn Mode

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-skill-generator/references/LEARN-MODE.md

These examples show how to invoke the AI Factory Skill Generator with URLs to activate Learn Mode. A non-URL argument can be used as a skill name hint.

```bash
/aif-skill-generator https://docs.example.com/api-reference
```

```bash
/aif-skill-generator https://react.dev/learn https://react.dev/reference
```

```bash
/aif-skill-generator my-skill-name https://some-docs.com/guide
```

```bash
/aif-skill-generator https://github.com/owner/repo https://blog.example.com/best-practices
```

--------------------------------

### Docker Compose Up Development Command

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-dockerize/references/SUMMARY-FORMAT.md

Starts the development environment using Docker Compose.

```bash
docker compose up
```

--------------------------------

### Generate README.md Structure

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-docs/SKILL.md

Provides a template structure for generating a README.md file, including sections for project name, tagline, description, quick start, key features, and an example.

```markdown
# Project Name

> One-line tagline describing the project.

Brief 2-3 sentence description of what this project does and why it exists.

## Quick Start

```bash
# Installation steps (1-3 commands)
```

## Key Features

- **Feature 1** — brief description
- **Feature 2** — brief description
- **Feature 3** — brief description

## Example

```
```

--------------------------------

### PHP CI Setup with Shared Configuration

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-ci/references/BEST-PRACTICES.md

Define a reusable PHP CI configuration using YAML anchors and `extends` in GitLab CI. Includes image, package installation, and Composer dependency caching.

```yaml
.php-setup:
  image: php:8.3-cli
  before_script:
    - apt-get update && apt-get install -y git unzip
    - curl -sS https://getcomposer.org/installer | php -- --install-dir=/usr/local/bin --filename=composer
    - composer install --no-interaction --prefer-dist
  cache:
    - key:
        files:
          - composer.lock
      paths:
        - vendor/
      policy: pull

phpstan:
  extends: .php-setup
  stage: lint
  script:
    - vendor/bin/phpstan analyse

```

--------------------------------

### npx Install Scan Workflow

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-skill-generator/references/SECURITY-SCANNING.md

Steps for scanning a skill after installation using `npx skills install`.

```text
1. npx skills install {{skills_cli_agent_flag}} <name>  # Downloads skill
2. LEVEL 1: If a Python 3 command was detected, run the automated scan on the installed directory with that concrete command, for example `python3 ~/{{skills_dir}}/aif-skill-generator/scripts/security-scan.py <installed-directory>`.
3. If Python 3 is unavailable, do not invoke `security-scan.py`; report "Level 1 skipped: Python 3 unavailable" and continue to Level 2 only if the user accepted that risk.
4. LEVEL 2: Read and review the skill content semantically
5. If BLOCKED and Python 3 is available → run the cleanup helper with the same selected Python 3 command, for example `python3 ~/{{skills_dir}}/aif-skill-generator/scripts/cleanup-blocked-skill.py --skill <name> --installed-path <installed-directory>` (reuse the same `<installed-directory>` you scanned in step 2 — upstream `skills` sanitizes the on-disk directory name, so synthesizing `{{skills_dir}}/<name>` can miss the real folder; `--installed-path` lets the helper verify physical removal) and warn user.
6. If BLOCKED and Python 3 is unavailable → do not use the skill; ask the user to remove the installed directory or provide Python 3 so the cleanup helper can run.
```

--------------------------------

### Install AI Factory via npm

Source: https://github.com/lee-to/ai-factory/blob/2.x/README.md

Installs the AI Factory CLI globally using npm. This is the primary method for installation.

```bash
npm install -g ai-factory
```

--------------------------------

### Install AI Factory via mise

Source: https://github.com/lee-to/ai-factory/blob/2.x/README.md

Installs the AI Factory CLI globally using mise, a task runner and installer. Useful for managing multiple tool versions.

```bash
mise use -g npm:ai-factory
```

--------------------------------

### Skill Structure Examples (File System)

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-skill-generator/references/BEST-PRACTICES.md

Illustrates file organization for skills of varying complexity, from small to large.

```text
skill-name/
└── SKILL.md
```

```text
skill-name/
├── SKILL.md
└── references/
    └── REFERENCE.md
```

```text
skill-name/
├── SKILL.md
├── references/
│   ├── REFERENCE.md
│   ├── API.md
│   └── EXAMPLES.md
├── scripts/
│   ├── main.py
│   └── helpers.py
├── templates/
│   └── output.md
└── assets/
    └── schema.json
```

--------------------------------

### Install and Scan External Skills

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif/SKILL.md

Installs an external skill using the skills CLI and immediately performs a security scan on the installed path. Handles different exit codes for blocking, warnings, or clean states.

```bash
npx skills install {{skills_cli_agent_flag}} <name>
# AUTO-SCAN: immediately after install. Example for PYTHON_CMD=(python3).
python3 ~/{{skills_dir}}/aif-skill-generator/scripts/security-scan.py <installed-path>
# Exit 1 (BLOCKED) → run cleanup script
# Exit 2 (WARNINGS) → show to user, ask confirmation
# Exit 0 (CLEAN) → read files yourself (Level 2), verify intent, proceed
```

--------------------------------

### Example Plan During Execution

Source: https://github.com/lee-to/ai-factory/blob/2.x/docs/subagents.md

Illustrates how the plan file is annotated during the execution of the implement-coordinator agent, showing task status and dependencies.

```markdown
### Phase 1: Setup
<!-- parallel: tasks 1, 2 -->
- [x] Task 1: Create User model
- [~] Task 2: Add authentication types <!-- in-progress -->

### Phase 2: Core
- [ ] Task 3: Implement password hashing (depends on 1, 2)
- [ ] Task 4: Create auth service (depends on 3)
```

--------------------------------

### Example Self-Improvement Patch

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-fix/SKILL.md

An example of a completed self-improvement patch for a null reference error in UserProfile.

```markdown
# Null reference in UserProfile when user has no avatar

**Date:** 2026-02-07 14:30
**Files:** src/components/UserProfile.tsx
**Severity:** medium

## Problem

TypeError: Cannot read property 'url' of undefined when rendering
UserProfile for users without an uploaded avatar.

## Root Cause

The `user.avatar` field is optional in the database schema but the
component accessed `user.avatar.url` without a null check. This was
introduced in commit abc123 when avatar display was added — the
developer tested only with users that had avatars.

## Solution

Added optional chaining: `user.avatar?.url` with a fallback to a
default avatar URL. Also added a null check in the Avatar sub-component.

## Prevention

- Always check if database fields marked as `nullable` / `optional`
  are handled with null checks in the UI layer
- Add test cases for "empty state" — user with minimal data
- Consider a lint rule for accessing nested optional properties

## Tags

`#null-check` `#react` `#optional-field` `#typescript`
```

--------------------------------

### Valid Name Field Examples

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-skill-generator/references/SPECIFICATION.md

Provides examples of valid naming conventions for the 'name' field in SKILL.md.

```text
pdf-processing
data-analysis
code-review
```

--------------------------------

### Next Steps for Full Mode (Normal)

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-plan/SKILL.md

Provides instructions after a plan is created in full mode without parallel execution. The user is prompted to review and manually start implementation.

```bash
Plan created with [N] tasks.
Plan file: <configured plans dir>/<resolved-plan-file>      # see Step 1.2 / Step 5 for the exact stem

To start implementation, run:
/aif-implement

To view tasks:
/tasks (or use TaskList)
```

--------------------------------

### Justfile: Parameters and Variadic Arguments

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-build-automation/references/BEST-PRACTICES.md

Shows how to define recipes with required parameters, optional parameters with defaults, and variadic arguments.

```justfile
# Required parameter
build target:
    go build -o bin/{{target}} ./cmd/{{target}}

# Optional parameter with default
deploy env="staging":
    ./deploy.sh {{env}}

# Variadic arguments
test *args:
    go test {{args}} ./...
```

--------------------------------

### Run Implement Coordinator for Implementation

Source: https://github.com/lee-to/ai-factory/blob/2.x/docs/subagents.md

Starts the implement coordinator agent to execute a pre-existing plan, reading the active plan or a specified file.

```bash
claude --agent implement-coordinator
```

```bash
claude --agent implement-coordinator "@.ai-factory/plans/feature-auth.md"
```

--------------------------------

### Mage Installation Command

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-build-automation/references/SUMMARY-FORMAT.md

Provides the command to install the Mage build tool using Go modules or Homebrew.

```bash
go install github.com/magefile/mage@latest
```

```bash
brew install mage
```

--------------------------------

### Justfile: Default Recipe for Help

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-build-automation/references/BEST-PRACTICES.md

Defines a default recipe that displays available tasks when Just is run without arguments, enhancing discoverability.

```justfile
[doc("Show available recipes")]
default:
    @just --list --unsorted
```

--------------------------------

### Conventional Commit Example: Simple Feature

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-commit/SKILL.md

Example of a simple feature commit using the conventional commit format.

```markdown
feat(auth): add password reset functionality
```

--------------------------------

### Tool Specification Syntax Examples

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-skill-generator/references/SPECIFICATION.md

Illustrates various ways to specify allowed tools, including basic lists, bash commands with patterns, and combined lists.

```yaml
# Basic tools
allowed-tools: Read Write Grep Glob
```

```yaml
# Bash with patterns
allowed-tools: Bash(git *) Bash(npm run *) Bash(docker *)
```

```yaml
# Combined
allowed-tools: Read Write Bash(git *) Bash(python scripts/*.py)
```

--------------------------------

### Conventional Commit Example: Breaking Change

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-commit/SKILL.md

Example of a commit that introduces a breaking change, indicated by '!' and a BREAKING CHANGE footer.

```markdown
feat(api)!: change response format for user endpoint

BREAKING CHANGE: user endpoint now returns nested profile object
```

--------------------------------

### Starting New Feature from Current Branch

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-implement/SKILL.md

Use this command to start a new feature plan based on the current branch.

```bash
/aif-plan full <description>
```

--------------------------------

### Install and Scan External Skill

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-skill-generator/SKILL.md

Installs an external skill and runs a security scan. If the scan is blocked, a cleanup script is provided.

```bash
npx skills install {{skills_cli_agent_flag}} <name>
python3 ~/{{skills_dir}}/aif-skill-generator/scripts/security-scan.py <installed-path>
```

```bash
python3 ~/{{skills_dir}}/aif-skill-generator/scripts/cleanup-blocked-skill.py --skill <name> --installed-path <installed-path>
```

--------------------------------

### Justfile: Cross-Platform Support with OS-Specific Recipes

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-build-automation/references/BEST-PRACTICES.md

Illustrates how to define OS-specific recipes using attributes like `[linux]`, `[macos]`, and `[windows]` for platform-dependent commands.

```justfile
# OS-specific commands
[linux]
install-deps:
    sudo apt-get install -y build-essential

[macos]
install-deps:
    brew install gcc

[windows]
install-deps:
    choco install mingw
```

--------------------------------

### Task Completion Status Examples

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-verify/SKILL.md

Examples of verification results for tasks, indicating their completion status (COMPLETE, PARTIAL, NOT FOUND, SKIPPED).

```text
✅ Task #1: Create user model — COMPLETE
   - User model created at src/models/user.ts
   - All fields present (id, email, name, createdAt, updatedAt)
   - Validation decorators added

⚠️ Task #3: Add password reset endpoint — PARTIAL
   - Endpoint created at src/api/auth/reset.ts
   - MISSING: Email sending logic (task mentioned SendGrid integration)
   - MISSING: Token expiration check

❌ Task #5: Add rate limiting — NOT FOUND
   - No rate limiting middleware detected
   - No rate-limit related packages in dependencies
```

--------------------------------

### Interactive Setup: Database Choice

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-dockerize/SKILL.md

Prompts the user to select the database technology for the project when generating new Docker configurations. This influences the generated Dockerfile and Compose services.

```bash
AskUserQuestion: Which database does this project use?

Options:
1. PostgreSQL (Recommended)
2. MySQL / MariaDB
3. MongoDB
4. SQLite (no container needed)
5. None
```

--------------------------------

### Conventional Commit Example: Bug Fix with Body

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-commit/SKILL.md

Example of a bug fix commit that includes a body explaining the issue and resolution.

```markdown
fix(api): handle null response from payment gateway

The payment API can return null when the gateway times out.
Added null check and retry logic.

Fixes #123
```

--------------------------------

### Load Project Context Description

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-build-automation/SKILL.md

Reads the project description file to store context like tech stack and framework. This step is optional; if the file doesn't exist, context will be detected later.

```bash
Read .ai-factory/DESCRIPTION.md
```

--------------------------------

### Quality Gate Pass Status Example

Source: https://github.com/lee-to/ai-factory/blob/2.x/docs/quality-gates.md

Example of a quality gate passing. This JSON structure indicates no blocking issues were found.

```aif-gate-result
{
  "schema_version": 1,
  "gate": "review",
  "status": "pass",
  "blocking": false,
  "blockers": [],
  "affected_files": ["skills/aif-review/SKILL.md"],
  "suggested_next": {
    "command": "/aif-commit",
    "reason": "Review found no blocking issues."
  }
}
```

--------------------------------

### Quality Gate Warn Status Example

Source: https://github.com/lee-to/ai-factory/blob/2.x/docs/quality-gates.md

Example of a quality gate returning a warning status. This indicates potential issues that do not block the process.

```aif-gate-result
{
  "schema_version": 1,
  "gate": "rules",
  "status": "warn",
  "blocking": false,
  "blockers": [],
  "affected_files": [],
  "suggested_next": {
    "command": "/aif-rules",
    "reason": "Rules are missing or ambiguous for the changed scope."
  }
}
```

--------------------------------

### Start AI Factory Implementation from Specific Task

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-implement/SKILL.md

Starts the implementation process from a specified task number, useful for skipping or re-doing tasks.

```bash
/aif-implement 5
```

--------------------------------

### Full Mode Scenario (Normal)

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-plan/references/EXAMPLES.md

Demonstrates the 'full' mode execution, including reconnaissance, branch creation, milestone linkage prompts, test/logging/docs questions, deep codebase exploration, and task creation with commit checkpoints.

```text
/aif-plan full Add user authentication with OAuth

-> mode=full
-> Quick reconnaissance
-> Plan slug: user-authentication
-> Branch: feature/user-authentication (if git branch creation is enabled)
-> If ROADMAP.md exists: asks about milestone linkage, user picks one (or skips)
-> Asks about tests (Yes), logging (Verbose), docs (Yes)
-> Creates branch only when `git.enabled=true` and `git.create_branches=true`
-> Explores codebase deeply
-> Creates 8 tasks with commit checkpoints
-> Saves plan to `paths.plans/feature-user-authentication.md` (or `paths.plans/user-authentication.md` when no branch is created)
-> When `workflow.plan_id_format = sequential`, the filename gains a 4-digit prefix:
   `paths.plans/0007_feature-user-authentication.md`
-> STOP - user runs /aif-implement when ready
```

--------------------------------

### Cleanup Blocked Skill

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif/SKILL.md

This command is executed when a skill installation fails security checks (exit code 1). It cleans up the partially installed skill.

```bash
~/{{skills_dir}}/aif-skill-generator/scripts/cleanup-blocked-skill.py --skill <name> --installed-path <installed-path>
```

--------------------------------

### System Dynamic Context Commands

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-skill-generator/templates/dynamic-context.md

Examples of system commands for checking software versions and system information, suitable for dynamic context integration. These commands help in understanding the execution environment.

```bash
node --version
```

```bash
python --version
```

```bash
uname -a
```

```bash
cat package.json | jq '.name, .version'
```

--------------------------------

### PHP Composer Install Commands

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-build-automation/references/BEST-PRACTICES.md

Commands for installing Composer dependencies. The first is CI-friendly and reproducible, while the second is optimized for production by skipping development dependencies.

```bash
# Install (CI-friendly, reproducible)
composer install --no-interaction --prefer-dist --optimize-autoloader
```

```bash
# Install for production (skip dev deps)
composer install --no-dev --no-interaction --prefer-dist --optimize-autoloader
```

--------------------------------

### Skill-Context File Format Example

Source: https://github.com/lee-to/ai-factory/blob/2.x/docs/evolve.md

Demonstrates the structure of a skill-context file, including metadata and auto-generated rules derived from project patches.

```markdown
# Project Rules for /aif-fix

> Auto-generated by `/aif-evolve`. Do not edit manually.
> Last updated: 2026-02-15 14:30
> Based on: 12 analyzed patches

## Rules

### Null-check for optional relations
**Source**: 2026-02-10-09.15.md, 2026-02-12-16.45.md, 2026-02-14-11.20.md
**Rule**: Check for optional/nullable fields before accessing nested
properties. This is the #1 source of bugs in this project — 5 of 12
patches involved null references on optional DB relations.

### Prisma unique constraint
**Source**: 2026-02-13-10.00.md
**Rule**: Before `create()` calls on models with unique fields, check
for existing records with `findUnique()`. Use upsert when appropriate.

```

--------------------------------

### Generate Git Branch Name Examples

Source: https://github.com/lee-to/ai-factory/blob/2.x/skills/aif-plan/SKILL.md

Examples of generated git branch names based on common development workflows. These follow a `type/description` format.

```text
feature/user-authentication
fix/cart-total-calculation
refactor/api-error-handling
chore/upgrade-dependencies
```