Support and detailed guides for this repo can be found here: https://www.skool.com/appbuilders

A repository-native prompt architecture system that helps you build features quickly with AI coding assistants (Cursor, Claude Code) without drifting, skipping steps, or rebuilding context every time.

This is a template repository that turns your codebase into a self-documenting AI operating system.

Modern AI coding tools are powerful, but they tend to drift, skip steps, and rebuild context every time. This system solves that by organizing AI knowledge into stable, repository-specific files that separate:

• System knowledge – how your app actually works
• Implementation patterns – how to do common tasks correctly
• Workflows – step-by-step processes for building features

The result: speed without degradation. Your AI assistant knows your codebase, follows your conventions, and executes systematically.

# Copy the entire .ai directory structure into your repo root
cp -r .ai /path/to/your/project/

In Cursor or Claude Code, run:

@.ai/workflows/install-system.prompt.md

This scans your codebase and populates the system knowledge files with your repository-specific information.

Audit your system readiness at any stage:

@.ai/workflows/system-readiness-audit.prompt.md

This analyzes your codebase and tells you exactly what system functionality you need for your current stage and scale.

Create a PRD:

@.ai/workflows/create-prd.prompt.md

Feature:
Add user profile pages with edit functionality

Execute the PRD:

@.ai/workflows/execute-prd.prompt.md
PRD: prd/01-user-profile-pages.md

The AI will implement step-by-step, following your codebase conventions and patterns.

The system uses three types of files:

1. System files (.ai/system/) - Describe how your repository works (stack, conventions, architecture)
2. Pattern files (.ai/patterns/) - Describe how to implement common tasks (migrations, APIs, navigation, etc.)
3. Workflow prompts (.ai/workflows/) - Executable instructions for complex processes (creating PRDs, executing features, auditing)

See .ai/README.md for detailed documentation.

• install-system.prompt.md - Bootstrap the system (run this first)
• system-readiness-audit.prompt.md - Audit your codebase to identify what system infrastructure you need at your current stage
• create-prd.prompt.md - Create a technical PRD for a feature
• execute-prd.prompt.md - Execute a PRD step-by-step

See .ai/workflows/README.md for full documentation on all workflows.

This template provides a foundation, but you can extend it:

• Add custom patterns in .ai/patterns/ for recurring tasks specific to your codebase
• Extend system guidelines in .ai/system/ with additional conventions and architecture details
• Customize workflows if needed, though they're designed to work for most projects out of the box

• .ai/README.md - Architecture overview
• .ai/system/README.md - System knowledge files
• .ai/patterns/README.md - Implementation patterns
• .ai/workflows/README.md - Workflow documentation

Note: This is a template. The files in system/ and patterns/ start as placeholders. Run the install workflow to populate them with your repository's actual information.