The-VibeCode-Bible

agent
Security Audit
Fail
Health Warn
  • No license — Repository has no license file
  • Description — Repository has a description
  • Active repo — Last push 0 days ago
  • Community trust — 36 GitHub stars
Code Fail
  • network request — Outbound network request in next.config.mjs
  • exec() — Shell command execution in src/app/[slug]/SectionClient.tsx
Permissions Pass
  • Permissions — No dangerous permissions requested

No AI report is available for this listing yet.

SUMMARY

A free, living knowledge base on vibe coding — building SaaS fast with AI agents, local AI, and agentic workflows.

README.md

⚡ The VibeCode Bible

Instagram @web.dev.george
Typing SVG

This repository is the personal knowledge base of @web.dev.george. It contains everything I know about building SaaS applications using AI, structured agent skills, and "vibe coding". No fluff.


🚀 Quick Start

  1. Clone the repository
git clone https://github.com/WebDevPeterGriffin/The-VibeCode-Bible.git
cd The-VibeCode-Bible
  1. Install dependencies
npm install
  1. Run the local server
npm run dev

Open http://localhost:3000 with your browser to explore the bible.


🧠 What's Inside?

📚 40+ Lessons across 4 Topic Areas

Foundation

  • My Exact Setup — the stack, extensions, folder structure, and settings I use daily
  • The AGENTS.md System — how I give AI agents permanent context so they stop making dumb mistakes
  • Antigravity vs Claude Code — when to use which one and why, from someone who uses both daily
  • The Vibe Coding Workflow — the full 7-step process from brain dump to production deploy

Building & Shipping

  • Real Project Walkthrough — how I built Siteo in 3 days using this exact workflow
  • Supabase RLS Patterns — Row Level Security done right, with AI generating every migration
  • Beating the Generic AI Look — skill files and design tokens that stop AI-generated aesthetics

Local AI & Agentic

  • Setting Up Local AI — installing Ollama, pulling models, and running local AI with zero API costs
  • Vibe Coding Is Dead — why pure vibe coding failed at scale and what replaced it
  • Agentic Engineering — you are the governor, the AI is the workforce
  • Agent Harness Engineering — the layer that controls what the model sees, what tools it uses, and when it stops. Covers the anatomy of a harness, building your first one from scratch, and running harnesses in production with observability and cost control
  • Fine-Tuning & LoRAs — a full 8-lesson deep dive on teaching local models new behaviour. Starts dead simple (what a LoRA even is, and how it differs from a skill file) and builds to training your own with Unsloth and running it in Ollama with your agent

Advanced & Scale

  • Claude Opus 4.7 Updates — extended thinking, new tokenizer, effort parameter, /ultrareview
  • Hermes Agent — the open-source self-improving agent with 156k GitHub stars. Runs 24/7 on a $5 VPS
  • Defending Against Prompt Injections — practical hardening techniques for local LLMs

🖥️ Built-in Local Chat

An interactive chat interface to safely test your prompts and skill files entirely locally. The assistant knows the full content map of the bible and links directly to lessons.

  • Auto-Model Discovery — Automatically detects your installed Ollama models (http://localhost:11434/api/tags)
  • Zero Cost, Total Privacy — Iterate on skill files totally offline and completely free
  • Prompt Injection Hardened — Built-in defenses via strict prompt boundaries and temperature lockdown
  • Multi-session — Full chat history with named sessions, persisted in localStorage

🌐 Extra Pages

  • MCP Servers (/mcp-servers) — Reference guide to Model Context Protocol servers and how to use them with AI agents
  • GEO + SEO (/geo-seo) — Generative Engine Optimization reference for AI-powered search visibility (ChatGPT, Perplexity, Google AI Overviews)

🛡️ Quick Security Audit

A dedicated sidebar panel that scans public GitHub repositories for hardcoded secrets, API keys, and database credentials using localized LLM analysis.

  • Secure by Default — runs completely locally via Ollama, ensuring code analysis doesn't leak sensitive config
  • Terminal UI — authentic audit reports with categorized severity (CRITICAL/HIGH/MEDIUM) levels

🤖 1,316 Bundled AI Agent Skills

Pre-loaded and ready to use the second you clone. No extra installs needed.

Skill Pack Files Original Creator What It Does
Awesome CursorRules 1,242 @PatrickJS Hundreds of framework-specific .cursorrules for Next.js, React, Vue, Python, and more
Claude SEO 68 @AgriciDaniel 13 sub-skills and 7 subagents for full technical SEO automation
GSAP Skills 33 @greensock GSAP/ScrollTrigger animation patterns, React/Next.js integration, and cinematic scroll effects
UI/UX Design Agent 31 @nextlevelbuilder Deep design intelligence — 50 styles, 21 palettes, 50 font pairings
Honnibal Claude Skills 10 @honnibal Mutation testing, type tightening, hypothesis tests, contract docstrings

⚡ 10 Executable Agent Workflows

Slash-command automation scripts your AI agent can run directly. The system automatically discovers new workflows added to .agent/workflows/.

Workflow Command Steps
Auth & Authorization /authentication-&-authorization 11
Code Quality Review /code-review 7
Debug Session /Debug-session 10
Deploy Checklist /deploy-checklist 8
Landing Page Generator /landing-page 7
New Feature Scaffold /new-feature 6
Performance Audit /Performance-audit 10
Project Cleanup /project-cleanup 8
Security Audit /security-audit 10
SEO Audit /seo-audit 7

🏗️ Production-Grade AGENTS.md

A battle-tested system prompt that enforces Next.js 14, TypeScript strict mode, Tailwind v4, dark-mode-first aesthetics, semantic HTML, and conventional commits across every AI interaction.


☕ Support

I built this and made it completely free. If it saved you time, helped you ship faster, or taught you something — consider leaving a donation. No pressure. Just vibes.

Donate
Built by @web.dev.george · Clone it. Run it. Read it.

Reviews (0)

No results found