GitHub - orangebread/wrinkl: Wrinkl is an AI context management system with ledger-based feature tracking for better AI-assisted development

5 min read Original article ↗

❗️⚠️ PROJECT DEPRECATED ⚠️❗️: Please try Wrinkl's successor SpecLinter https://github.com/orangebread/speclinter-mcp. Thanks to all who starred and supported this project!

Wrinkl Logo

🧠 AI Context Management System

A context management system for AI-assisted development ✨

Track features with ledgers 📋 • Maintain coding patterns 🎯 • Keep AI assistants aligned with your project 🤖

📦 Installation

Choose your preferred package manager:

# npm
npm install -g wrinkl

# pnpm
pnpm add -g wrinkl

# yarn
yarn global add wrinkl

🚀 Quick Start

# Initialize in your project
cd my-project
wrinkl init

# Create a feature ledger
wrinkl feature user-authentication

# List active features
wrinkl list

# Archive completed features
wrinkl archive user-authentication

🔥 ULTIMATE PROTIP: After running wrinkl init, ask your AI coding assistant to automatically populate the entire .ai/ directory for you! ✨

🎯 What It Does

Wrinkl creates a .ai/ directory in your project with:

  • 📄 Context files for AI assistants to understand your project
  • 📚 Pattern documentation to maintain consistency
  • 📋 Feature ledgers to track work progress
  • 🏗️ Architecture decisions to guide development

💡 Why Use This?

  • 🤖 Better AI Assistance - AI tools understand your project context
  • 📝 Feature Tracking - Ledgers document progress and decisions
  • 🎯 Pattern Consistency - Maintain coding standards across the team
  • 🔄 Living Documentation - Context evolves with your project

Important: Keep your feedback loops tight! AI works better on focused tasks rather than sprawling features

🚀 The Story Behind Wrinkl

"After 2+ years of coding exclusively with AI, I've learned that context is everything."

As a software engineer with 15 years of experience, I've witnessed the AI revolution transform how we build software. Wrinkl is my attempt to formalize the patterns and processes that make AI-assisted development truly effective.

The Problem: AI assistants are incredibly powerful, but they often lack the context needed to make decisions that align with your project's goals, patterns, and constraints.

The Solution: A structured approach to context management that keeps your AI assistants informed, your team aligned, and your codebase consistent.

This isn't just another tool—it's a methodology that evolves with the rapidly changing AI landscape.

💬 Want to chat about AI-assisted development? Hit me up on Discord: jayeeeffeff

📁 Directory Structure

After running wrinkl init, you'll have:

your-project/
├── .ai/
│   ├── README.md              # Overview of the AI context system
│   ├── project.md             # Project overview and requirements
│   ├── patterns.md            # Coding patterns and conventions
│   ├── architecture.md        # System architecture and decisions
│   ├── context-rules.md       # Rules for AI assistants
│   └── ledgers/
│       ├── _active.md         # Dashboard of active features
│       ├── _template.md       # Template for new feature ledgers
│       ├── archived/          # Completed feature ledgers
│       └── [feature-name].md  # Individual feature ledgers
├── .cursorrules               # Cursor AI rules (optional)
├── augment.md                 # Augment AI context (optional)
└── .github/
    └── copilot-instructions.md # GitHub Copilot instructions (optional)

⚡ Commands

🎬 wrinkl init

Initialize the AI context system in your project.

Options:

  • -n, --name <name> - Project name (default: directory name)
  • -t, --type <type> - Project type (default: "web app")
  • -s, --stack <stack> - Technology stack (default: "TypeScript, Node.js")
  • --no-cursor - Skip creating .cursorrules file
  • --with-augment - Include augment.md file
  • --with-copilot - Include GitHub Copilot instructions

Example:

wrinkl init --name "My App" --type "mobile app" --stack "React Native, Node.js"

🆕 wrinkl feature <name>

Create a new feature ledger to track development progress.

Example:

wrinkl feature user-authentication

📋 wrinkl list

List all active feature ledgers and their status.

Options:

  • -a, --all - Include archived features

Example:

📦 wrinkl archive <name>

Archive a completed feature ledger.

Example:

wrinkl archive user-authentication

⚙️ How It Works

1. 📊 Project Context

The .ai/project.md file contains your project's core information:

  • Project goals and constraints
  • Technology stack
  • Key requirements
  • Development workflow

2. 🎨 Coding Patterns

The .ai/patterns.md file documents:

  • Code style and conventions
  • Common patterns and anti-patterns
  • Testing strategies
  • Performance guidelines

3. 🏗️ Architecture Decisions

The .ai/architecture.md file captures:

  • System design decisions
  • Technology choices and trade-offs
  • Scalability considerations
  • Security architecture

4. 📋 Feature Ledgers

Individual feature files track:

  • Feature requirements and goals
  • Technical approach and decisions
  • Progress updates and blockers
  • Testing and deployment notes

5. 🤖 AI Assistant Rules

The .ai/context-rules.md file provides:

  • Guidelines for AI assistants
  • Code quality standards
  • Security and performance rules
  • Project-specific requirements

🌟 Best Practices

👥 For Teams

  1. Keep context updated 🔄 - Regularly update project files as requirements change
  2. Use feature ledgers 📝 - Create a ledger for each significant feature
  3. Document decisions 📋 - Record important technical decisions in ledgers
  4. Review patterns 🔍 - Regularly review and update coding patterns

🤖 For AI Assistance

  1. Reference context 📖 - Tell AI assistants to read the .ai/ directory
  2. Mention features 🎯 - Reference specific feature ledgers when working
  3. Update progress ⏱️ - Keep ledgers updated with progress and decisions
  4. Follow patterns ✅ - Ensure AI-generated code follows project patterns

💬 Example AI Prompts

"I'm working on the user-authentication feature. Please read the feature
ledger in .ai/ledgers/user-authentication.md and help me implement the
login component following the patterns in .ai/patterns.md"

"Please review the project context in .ai/project.md and suggest an
architecture for the new notification system, documenting your decisions
in a new feature ledger"

🔗 Integration with AI Tools

🎯 Cursor AI

If you use Cursor, the .cursorrules file provides context and guidelines for the AI assistant.

🌊 Windsurf AI

If you use Windsurf, simply rename .cursorrules to .windsurfrules - the content is identical, just different filename conventions.

⚡ Augment AI

The augment.md file provides context for Augment AI when working on your project.

🐙 GitHub Copilot

The .github/copilot-instructions.md file guides GitHub Copilot to generate code that follows your project patterns.

🤝 Contributing

  1. 🍴 Fork the repository
  2. 🌿 Create a feature branch
  3. ✏️ Make your changes
  4. 🧪 Add tests
  5. 📤 Submit a pull request

📄 License

MIT - see LICENSE file for details.

🆘 Support