Vibe Code Bible

Vibe Code Bible

The open-source knowledge base for vibe coders. Feed these `.md` files to your AI coding assistant to ship faster, smarter, and with best practices baked in.

Vibe Code Bible

The open-source knowledge base for vibe coders. Feed these .md files to your AI coding assistant to ship faster, smarter, and with best practices baked in.


What Is This?

Vibe coding is building software with AI as your copilot. The bottleneck isn't writing code anymore — it's giving your AI the right context. That's what this repo solves.

Each .md file is a self-contained knowledge module covering principles, LLM-ready instructions, and real-world examples. Copy any file into your AI session and your assistant instantly becomes a domain expert.

How to Use

1. Find the topic you need       → Browse the folders below
2. Copy the .md file contents    → Paste into your AI coding session
3. Stack multiple files           → Combine topics for complex tasks
4. Build with confidence         → Your AI now has expert-level context

Example workflow:

  • Building an e-commerce site? Feed your AI Product-Page-SEO/product-page-seo.md + Structured-Data/structured-data.md + Accessibility/accessibility.md
  • Designing a dashboard? Feed it Design-Systems/design-systems.md + Responsive-Design/responsive-design.md + Data-Visualization/data-visualization.md

Chapters

ChapterTopicsStatus
SEOTechnical SEO, On-Page, Off-Page, Content, Local, E-commerce, International, Enterprise, AI SEO, AnalyticsComplete
UI/UXBrand Identity, Design Systems, Accessibility, Responsive Design, Typography, UX Patterns, Animation, Dark Mode, Mobile-FirstComplete
FrontendReact 19, Next.js 15, CSS Architecture, Performance, State Management, Component Patterns, TypeScript, Data Fetching, FormsComplete
BackendAPI Design, Database Design, Caching, Error Handling, Auth & Sessions, Serverless, Background Jobs, Webhooks, Real-TimeComplete
DevOpsCI/CD, Docker, Cloud Architecture, Monitoring, Infrastructure as CodeComplete
SecuritySecrets Management, Authentication, Frontend Security, Backend Security, API Security, Data Protection, Security Headers, Supply Chain, Security TestingComplete
AI IntegrationLLM Patterns, Prompt Engineering, RAG, AI Agents, EmbeddingsComplete
TestingUnit Testing, E2E Testing, Performance Testing, Test StrategyPlanned
Product & GrowthAnalytics, Experimentation, Conversion Optimization, Onboarding, Retention, Email/Notifications, Referrals, Billing, Marketing Channels, Product-Led GrowthComplete
CopywritingSEO Copywriting, UX Writing, Landing Pages, Email Copy, CTAs & Conversion, Headlines & Hooks, Brand Voice & Tone, Product Copy, Content WritingComplete
ToolsBaaS Platforms, Authentication, Databases, Email Services, Analytics & Monitoring, Payments, Live Chat, Background Jobs, Hosting, Search, CMS, File Storage, Communication (~40 tools)Complete

File Format

Every file follows the same structure so your AI knows what to expect:

# Topic Name
> One-line summary

## Principles        → The "why" and core knowledge
## LLM Instructions  → Direct instructions your AI can follow
## Examples          → Real code, configs, and templates
## Common Mistakes   → What NOT to do

Principles

  • Self-contained — Every file works on its own. No dependencies.
  • LLM-optimized — Structured for AI consumption with clear, direct instructions.
  • Actionable — Real examples, real code, real configs. Not theory essays.
  • Opinionated — Best practices, not every possible option. We pick what works.
  • Current — Every file has a "Last reviewed" date. Outdated advice is worse than no advice.

Contributing

  1. Follow the Principles → LLM Instructions → Examples → Common Mistakes format
  2. Include code snippets where applicable (HTML, CSS, JS, config files, etc.)
  3. Be opinionated — recommend the best approach, not all approaches
  4. Add a Last reviewed: YYYY-MM date at the bottom
  5. Add See also: cross-references to related files
  6. Keep it practical — if a vibe coder can't use it in a session, it doesn't belong

License

MIT — Use it, fork it, build with it.


By Ryan Lind, Assisted by Claude Code and Google Gemini.

On this page