Build enterprise-ready brand systems with AI.
brand-manual-app is an open-source Brand OS that helps teams transform fragmented brand assets into a complete, governed design system across product, marketing, and customer experience.
- Runs structured brand onboarding.
- Applies curated identity archetypes (or full manual mode).
- Generates outputs across 55 modules in 7 phases.
- Lets teams edit module content in dedicated workspaces.
- Exports manuals to
PDF,HTML, andMarkdown. - Supports token-based public read-only sharing.
This is not “another UI kit repo.”
It is a full system that connects:
- brand strategy,
- design language,
- operational guidelines,
- and production-ready documentation.
If you want to contribute to a real-world AI + design-system platform with enterprise scope, this repo has meaningful work across architecture, UX, data modeling, AI contracts, and reliability.
Identity profiles are defined in src/lib/data/personalities.ts and include:
- philosophy,
- best-fit and non-fit contexts,
- typography DNA,
- color DNA,
- layout and motion rules,
- photography direction,
- voice and tone behavior.
Current profiles:
- Apple
- Teenage Engineering
- Polestar
- Linear
- Bang & Olufsen
- Nothing
- Aesop
- Manual mode
Deep dive: docs/IDENTITY_ARCHETYPES.md
Every generated brand system should be:
- Distinctive: clearly recognizable.
- Consistent: reusable across channels and teams.
- Scalable: useful from early-stage to enterprise growth.
- Operational: directly executable by product/marketing/CX teams.
- Governable: auditable and maintainable over time.
- Routes and APIs:
src/app/** - Workspaces and UI:
src/components/** - Domain data (phases/modules/identities):
src/lib/data/** - AI orchestration:
src/lib/ai/**,src/actions/** - Supabase integration:
src/lib/supabase/** - Export engines:
src/lib/exporters/**
- Next.js 16
- React 19
- TypeScript 5
- Tailwind CSS 4
- Supabase
- BAML
- OpenAI
git clone https://github.com/admin-blip7/desing-system.git
cd desing-system
npm install
cp .env.example .env.local
npm run devOpen: http://localhost:3000
Use .env.example as your template.
Minimum keys to run text generation:
NEXT_PUBLIC_SUPABASE_URL=your_supabase_project_url
NEXT_PUBLIC_SUPABASE_ANON_KEY=your_supabase_anon_key
OPENAI_API_KEY=your_openai_api_key
OPENAI_TEXT_MODEL=gpt-5-mini
AI_PROVIDER=openaiOptional keys for image generation:
GOOGLE_API_KEY=your_google_api_key_here
# GEMINI_API_KEY=your_google_api_key_hereYou do not need to publish anything to npm to use or contribute to this project.
Why:
- This repository is an application monorepo, not a public package template.
- Root
package.jsonis marked"private": true. - Internal workspace packages are local/private and resolved via npm workspaces.
For contributors, the correct setup is always:
- clone repo,
npm install,- configure
.env.local, - run
npm run dev.
npm run dev- start local development servernpm run build- production buildnpm run start- run production build locallynpm run lint- lint projectnpm run test:ai- AI package testsnpm run docs:ai- regenerate AI module docs
- Contribution guide: CONTRIBUTING.md
- Community roadmap: COMMUNITY_ROADMAP.md
- Execution playbook: docs/COMMUNITY_PLAYBOOK.md
Issue/PR templates are included under .github/.
- Never commit
.env.localor real secrets. - Never commit local tool artifacts or private dumps.
- Keep production error responses sanitized.
- Report potential secret exposure immediately.
Core flows are functional.
Primary community opportunity is finishing phase-level quality and production hardening.