Goal Kit transforms software development from task execution to outcome achievement using Goal-Driven Development methodology.

uv tool install --from git+https://github.com/Nom-nom-hub/goal-kit.git goalkeeper-cli

goalkeeper init my-project
cd my-project

Core workflow (vision → goal → strategy → milestones → execution):

/goalkit.vision          # Establish vision and principles
/goalkit.goal            # Define measurable goals
/goalkit.strategies      # Explore multiple strategies
/goalkit.milestones      # Create measurable milestones
/goalkit.execute         # Execute with learning

Extended workflow (execution → tasks → metrics → review):

/goalkit.tasks           # Break down into detailed tasks
/goalkit.report          # Generate progress reports
/goalkit.review          # Conduct retrospectives
/goalkit.taskstoissues   # Convert tasks to GitHub issues

Done! Your workflow is set up.

For the 5-minute walkthrough, see Quick Start Guide.

Goal-Driven Development focuses on outcomes over specifications:

Core Workflow (Always Use):

Extended Workflow (As Needed):

After goalkeeper init:

my-project/
├── .goalkit/
│   ├── vision.md                  # Project vision
│   └── goals/
│       └── 001-goal-name/
│           ├── goal.md            # Goal definition
│           ├── strategies.md       # Implementation approaches
│           ├── milestones.md       # Progress checkpoints
│           └── execution.md        # Implementation plan
├── CLAUDE.md                       # Agent context
├── CURSOR.md                       # Agent context
└── ... (your code)

For smaller projects or quick iterations, you can use the Lite templates:

• templates/lite-goal-template.md
• templates/lite-metrics-template.md

Works with all major AI coding assistants:

• Claude Code
• GitHub Copilot
• Google Gemini
• Cursor
• Qwen Code
• Windsurf
• Kilo Code
• Amazon Q
• opencode
• And others

From GitHub:

uv tool install --from git+https://github.com/Nom-nom-hub/goal-kit.git goalkeeper-cli

From local repo:

uv tool install --from . goalkeeper-cli

pip install git+https://github.com/Nom-nom-hub/goal-kit.git

Or locally:

pip install -e .

uv run --from git+https://github.com/Nom-nom-hub/goal-kit.git goalkeeper-cli init my-project

For detailed installation instructions, see Installation Guide.

• Quick Start Guide - 5-minute getting started
• Quick Reference Card - 📋 One-page cheat sheet (print and keep handy!)
• Installation Guide - Detailed install instructions for all platforms
• Agent & VSCode Setup - Setup AI agents and development environment

• Goal-Driven Development - Complete methodology guide
• Workflow Guide - 🔄 State machine, decision trees, and workflow patterns
• Common Mistakes - ❌ Avoid these pitfalls when starting
• Comparison with Spec-Driven - Key differences explained
• Practical Examples - Real-world use cases

• Troubleshooting Guide - Solutions for common issues
• Agent File Guide - Guide for AI agents using Goal Kit

• Changelog - Version history and release notes

Focus on outcomes, not implementation details

Always explore multiple valid approaches

Define clear metrics before building

Be willing to pivot based on evidence

Treat implementation as hypothesis testing

1. Define Vision
   ↓
2. Create Goal (with success metrics)
   ↓
3. Explore Strategies (3+ approaches)
   ↓
4. Plan Milestones (measurable steps)
   ↓
5. Execute (with continuous learning)
   ↓
6. Measure Results
   ↓
7. Repeat for next goal

• Python: 3.8+
• Git: For version control
• OS: Linux, macOS, or Windows
• uv: For package management (optional but recommended)

• Issues: Report on GitHub
• Questions: Ask in Discussions
• Troubleshooting: Read Troubleshooting Guide

• ✅ 9 commands for complete workflow (vision → goal → strategies → milestones → execute → tasks → report → review)
• ✅ Works with all major AI agents (Claude, Copilot, Cursor, Gemini, etc.)
• ✅ Cross-platform (Linux, macOS, Windows)
• ✅ Git integrated for branch management
• ✅ Measurable outcomes over tasks
• ✅ Multiple strategy exploration built-in
• ✅ Learning-focused execution
• ✅ Easy installation with uv
• ✅ VSCode optimized with settings templates and recommended extensions
• ✅ Agent integration guides included in box

Goal Kit includes built-in support for AI agents like Claude, Cursor, and Copilot:

1. Copy agent template: Use templates/agent-file-template.md to create CLAUDE.md (or other agents)
2. Share with agent: Provide the customized agent file when asking for Goal Kit help
3. Agent executes: The agent follows the methodology guides automatically
4. VSCode integration: Developers review and refine with optimized editor settings

See Agent & VSCode Setup for complete instructions.

Goal Kit provides optimized VSCode configuration:

• Auto-formatting for Python (Black), Markdown, JSON
• 100-character ruler for consistent code style
• GitLens integration for tracking goal milestones
• Recommended extensions for development (15+ tools)
• File nesting shows goal document relationships in Explorer

Setup in 2 steps:

cp templates/.vscode-settings-template.json .vscode/settings.json
cp templates/.vscode-extensions-template.json .vscode/extensions.json

VSCode will automatically suggest installing recommended extensions.

1. Install: uv tool install --from . goalkeeper-cli
2. Initialize: goalkeeper init my-project
3. Get Started: Read Quick Start Guide
4. Learn: Read Goal-Driven Development

Ready to focus on outcomes instead of specifications? Start with the Quick Start Guide.