github/spec-kit

💫 Toolkit to help you get started with Spec-Driven Development

What it solves

Spec Kit addresses the gap between high-level requirements and final code when using AI coding agents. It prevents agents from jumping straight to implementation—which often leads to bugs or misalignment—by enforcing a structured, spec-driven process where specifications are executable and directly generate the implementation.

How it works

Spec Kit provides a CLI tool (specify-cli) that integrates with over 30 AI coding agents (like GitHub Copilot or Claude). It introduces a series of structured slash commands that guide the agent through a specific development lifecycle:

  1. Constitution: Establish project-wide governing principles.
  2. Specify: Define the "what" and "why" (requirements).
  3. Plan: Define the technical architecture and tech stack.
  4. Tasks: Break the plan into actionable items.
  5. Implement: Execute the tasks to build the feature.
  6. Converge: Validate the code against the spec and plan, adding remaining work as new tasks.

It is highly extensible via extensions (adding new capabilities/commands) and presets (customizing how existing workflows behave).

Who it’s for

Software developers and organizations using AI coding agents who want to improve software quality, maintain consistency across teams, and ensure that AI-generated code strictly adheres to documented requirements.

Highlights

  • Agent Agnostic: Works with 30+ CLI and IDE-based AI coding agents.
  • Spec-Driven Development (SDD): Flips the script by making specifications the primary executable driver of the code.
  • Modular Customization: Supports extensions for new features, presets for workflow overrides, and bundles for role-based team setups.
  • Specialized Workflows: Includes opt-in extensions for evidence-based bug fixing (assess → fix → test) and idea assessment (intake → research → decide).

Related

  • Project
    loulanyue/spec-kit-zh

    A toolkit for Spec-Driven Development that provides structured prompt templates and a CLI to guide AI agents through a rigorous process of specification, planning, and implementation.

  • Project
    Priivacy-ai/spec-kitty

    A CLI for spec-driven development that provides a governed workflow and Git worktree isolation for AI coding agents.

  • Project
    clawplays/ospec

    A spec-driven, agentic workflow framework for AI coding agents that persists requirements, plans, and verification evidence directly in the project repository.

  • Project
    zhu1090093659/spec_driven_develop

    An architecture-first workflow plugin for AI coding agents that turns large software changes into a spec-driven loop of analysis, decomposition, and adaptive execution.

  • Dispatch
    Optimizing AI Coding Agents with Spec-Driven Development

    Explore a new Spec-Driven Development workflow for Claude Code that emphasizes decomposition, context management, and persistent specs to improve agent performance and reduce costs.