Maistik Spec KitSpec-Driven Development for teams that ship.

Specifications as source of truth. AI-assisted implementation. GitLab-native.

📝

Specs First, Code Second

Define what you want to build and why. Your AI assistant generates implementation from structured specifications — not the other way around.

🎨

Branding & Design Specs

Define colors, typography, spacing, and design tokens as structured specs. UI components reference these tokens for visual consistency.

🏗️

Infrastructure as Spec

Plan Docker Compose for development and Kubernetes/Helm for production. The infra spec is the source of truth for all deployment decisions.

🔄

Self-Improving Specs

Specs get smarter over time. The refinement engine detects gaps, staleness, and inconsistencies. Constitution updates propagate automatically.

🦊

GitLab-Native

Built for GitLab workflows. Uses GitLab API for templates, glab CLI for issues, and generates .gitlab-ci.yml pipelines.

🤖

AI-Powered with Claude

Works with Claude Code out of the box. Also supports 20+ other AI coding agents including Cursor, Gemini, Codex, and more.

The Workflow

Maistik Spec Kit follows a structured pipeline where each step produces a Markdown artifact that feeds the next:

Every artifact is a Markdown file versioned in Git. Specs are living documents that improve as your project grows.

Quick Example

bash

# Install
uv tool install maistik-specify-cli --from git+ssh://git@gitlab.com/maistik-studio/tools/spec-kit.git

# Initialize a project
specify init my-app --ai claude

# Then in Claude Code:
/speckit.constitution Code quality, testing, performance
/speckit.specify Build a task management app with kanban boards
/speckit.plan Use Next.js, PostgreSQL, and Tailwind CSS
/speckit.tasks
/speckit.implement

Get Started →