Spec Kit: The Complete Guide to GitHub's Official Spec-Driven Development Toolkit
Spec Kit is GitHub's official toolkit for Spec-Driven Development (SDD) — where specifications become executable, directly generating working implementations. A 6-step workflow from project principles to running code: /speckit.constitution → /speckit.specify → /speckit.plan → /speckit.tasks → /speckit.implement. 75,200+ stars, Python, MIT.
What Is Spec-Driven Development?
Spec-Driven Development flips the script on traditional software development. For decades, code was king — specifications were scaffolding, built and discarded once the "real work" of coding began. SDD changes this: specifications become executable, directly generating working implementations rather than just guiding them.
- Stars: 75,200+ ⭐
- Forks: 6,409
- Releases: 112
- Contributors: 103
- Watchers: 478
- Language: Python
- License: MIT
- Author: GitHub (Official)
- Topics: ai, copilot, spec, spec-driven, prd, engineering
The 6-Step Workflow
Step 1: Install Specify CLI
uv tool install specify-cli --from git+https://github.com/github/spec-kit.git
Step 2: Establish Project Principles
/speckit.constitution Create principles focused on code quality,
testing standards, user experience, and performance
Creates your project's governing principles — the guardrails that guide all development.
Step 3: Create the Spec
/speckit.specify Build an app that organizes photos into albums.
Albums grouped by date, drag-and-drop, tile preview.
Focus on the what and why, not the tech stack. Describe intent, not implementation.
Step 4: Create Implementation Plan
/speckit.plan The app uses Vite with vanilla HTML/CSS/JS.
Images stored locally, metadata in SQLite.
Now you provide tech stack and architecture choices.
Step 5: Break Down into Tasks
/speckit.tasks
Creates an actionable task list from your implementation plan.
Step 6: Execute Implementation
/speckit.implement
Executes all tasks and builds the feature according to plan.
Core Philosophy
| Principle | Description |
|---|---|
| Intent-Driven | Specifications define the "what" before the "how" |
| Rich Specs | Created using guardrails and organizational principles |
| Multi-Step Refinement | Not one-shot code generation from prompts |
| AI-Powered | Heavy reliance on advanced AI model capabilities |
Slash Commands Reference
| Command | Purpose |
|---|---|
/speckit.constitution | Create project principles and guidelines |
/speckit.specify | Describe what to build (functional spec) |
/speckit.plan | Define tech stack and architecture |
/speckit.tasks | Generate actionable task breakdown |
/speckit.implement | Execute implementation |
Why SDD Matters
Traditional AI coding:
Prompt → Code (one-shot, fragile)
Spec-Driven Development:
Principles → Specification → Plan → Tasks → Implementation
(multi-step, structured, reproducible)
The difference: SDD produces production-ready applications through structured refinement, not one-shot prompts. Each step builds on the previous, creating a chain of increasingly detailed specifications that the AI can interpret with precision.
Spec Kit vs Alternatives
Category: This is a spec-driven development toolkit for AI coding.
| Feature | Spec Kit | OpenSpec | Kiro | Cursor |
|---|---|---|---|---|
| Focus | SDD Toolkit | Spec format | AWS IDE | AI code editor |
| Stars | 75.2K ⭐ | ~2K ⭐ | N/A | N/A (commercial) |
| Author | GitHub (Official) | Fission AI | AWS | Anysphere |
| CLI Tool | ✅ Specify CLI | ❌ | ❌ | ❌ |
| Slash Commands | ✅ 5 commands | ✅ /opsx | ❌ | ❌ |
| Constitution/Principles | ✅ | ❌ | ❌ | Rules |
| Multi-Step Refinement | ✅ 6 steps | ✅ | ✅ | ❌ One-shot |
| Task Breakdown | ✅ | ❌ | ✅ | ❌ |
| Auto Implementation | ✅ | ❌ | ✅ | ✅ |
| Tech Stack Agnostic | ✅ | ✅ | ❌ | ✅ |
| Agent Support | ✅ Multiple | ✅ 20+ | Claude only | Built-in |
| Open Source | ✅ MIT | ✅ | ❌ | ❌ |
When to choose Spec Kit: You want GitHub's official SDD methodology — structured spec-to-code pipeline with multi-step refinement and CLI tooling.
When to choose OpenSpec: You want a spec format standard that works across 20+ AI tools.
When to choose Kiro: You want an AWS-backed IDE with built-in spec-driven features.
When to choose Cursor: You want a fast AI code editor for direct coding assistance.
Conclusion
Spec Kit is the most starred development methodology tool on GitHub with 75.2K stars — and it's from GitHub itself. It formalizes what experienced engineers have always known: good specifications lead to good software. By making specs executable through AI, Spec Kit turns this principle into a practical, repeatable workflow. The 6-step pipeline from constitution to implementation ensures that AI-generated code is structured, intentional, and production-ready.