AntiVibe

Understand any code, not just accept it.
A code learning & audit framework for Claude Code that turns any codebase — new, legacy, or AI-generated — into educational deep dives or senior-level architectural audits.

---

✨ What is AntiVibe?

AntiVibe is a code learning & audit framework that transforms any code — AI-generated, legacy, or otherwise — into educational content or architectural audits. You don't need recent git history or AI-authored files; point it at any file, directory, or module. Unlike generic code summaries, AntiVibe helps you understand:

• What the code does (functionality)
• Why it was written this way (design decisions)
• When to use these patterns (context)
• What alternatives exist (broader knowledge)

⚡ The Problem: AI writes code, developers copy-paste it, nobody learns anything.

🛡️ The Solution: AntiVibe explains the reasoning so you actually understand — and for code you already know, it audits the trade-offs instead.

---

🎯 Features

Feature	Description
Deep Dives	Generate comprehensive learning guides from any code
Audit Mode	Senior-level architectural audit — decisions, flags, edge cases, testability
Skill Levels	Tune depth to junior, mid, or senior
Output Modes	compact (default, low token cost) or full (resources + line-by-line)
Known Concepts	Skip list so familiar concepts get a one-line note, not a full explanation
Prerequisites	Maps each concept to the foundations you need to understand it first
Legacy-Friendly	Works on existing codebases — no git history or AI authorship required
Concept Mapping	Connect code to underlying CS principles
Curated Resources	Quality links to docs, tutorials, videos
Auto-Trigger	Optional hooks for automatic generation
Multi-Language	Works with any language/framework

---

🚀 Quick Start

Installation

# Clone the repository
git clone https://github.com/mohi-devhub/antivibe.git

# Install as a global Claude Code skill
cp -r antivibe ~/.claude/skills/antivibe

Usage

/antivibe                        # Start a deep dive (compact, mid-level by default)
/antivibe full                   # Full deep dive with resources + line-by-line
"deep dive"                      # Analyze recently written code (git diff)
"walk me through src/auth/"      # Explain an explicit file or directory
"explain this codebase"          # Understand existing / legacy code
"learn from this code"           # Generate learning guide

Tune the depth and detail inline:

"explain for a junior"           # Define terms, analogies, full snippets
"I know the basics"              # Mid level — focus on design decisions
"audit this, just the trade-offs"  # Senior mode — routes to the auditor agent

---

📁 Output Example

Run a full deep dive (/antivibe full) and get a file like the one below. The default compact mode produces a shorter version — Overview, Key Components, and Concepts (what + why) — without the walkthrough or resources.

# Deep Dive: Authentication System

## Overview
This auth system uses JWT tokens with refresh token rotation...

## Code Walkthrough
### auth/service.ts
- **Purpose**: Token generation and validation
- **Key Components**: 
  - `generateTokens()`: Creates access/refresh tokens
  - `verifyToken()`: Validates JWT signatures

## Concepts Explained
### JWT (JSON Web Tokens)
- **What**: Stateless authentication tokens...
- **Why**: Server doesn't need to store sessions...
- **When**: APIs, SPAs, microservices...

## Learning Resources
- [JWT.io](https://jwt.io): Official documentation
- [Auth0 Guide](https://auth0.com/blog): Best practices

Saved to: deep-dive/auth-system-2026-04-10.md

---

🔧 Configuration

Behavior (in SKILL.md)

Tune how AntiVibe analyzes code by editing the config blocks in SKILL.md:

output_mode: compact      # compact (default) or full
default_level: mid        # junior, mid, or senior
known_concepts:           # skipped in explanations (one-line note only)
  - async/await
  - React hooks
  - REST APIs

Setting	Options	Effect
output_mode	compact / full	compact keeps token cost low (no resources, no line-by-line); full adds curated resources, prerequisites, and a line-by-line walkthrough
default_level	junior / mid / senior	Sets explanation depth; senior routes to the auditor agent for an architectural audit
known_concepts	list	Concepts you already know — acknowledged in one line instead of fully explained

All three can also be overridden inline in a request (e.g. "/antivibe full", "explain for a junior", "just the trade-offs").

Auto-Trigger Hooks

Enable automatic deep-dive generation after task completion:

# Copy hooks to your project
cp hooks/hooks.json your-project/.claude/hooks.json

Hook	When	Use Case
SubagentStop	Task completes	Phase-based learning
Stop	Session ends	End-of-session summary

Customize Output Directory

Edit scripts/generate-deep-dive.sh:

OUTPUT_DIR="your-folder"  # Default: "deep-dive"

---

📂 File Structure

antivibe/
├── SKILL.md                     # Main skill definition
├── README.md                    # This file
├── hooks/
│   └── hooks.json              # Auto-trigger configuration
├── scripts/
│   ├── capture-phase.sh        # Detect implementation phases
│   ├── analyze-code.sh         # Parse code structure
│   ├── find-resources.sh       # Find external resources
│   └── generate-deep-dive.sh   # Generate markdown output
├── agents/
│   ├── explainer.md            # Subagent for learning-focused explanations
│   └── auditor.md              # Subagent for senior-level architectural audits
├── templates/
│   └── deep-dive.md            # Output template
├── reference/
│   ├── language-patterns.md    # Framework-specific patterns
│   └── resource-curation.md    # Curated learning resources
└── docs/
    ├── PLAN.md                  # Planning document
    └── setup.md                 # Detailed setup guide

---

📚 Principles

1. Why over what - Always explain design decisions
2. Context matters - Explain when/why to use patterns
3. Curated resources - Quality links, not random results
4. Phase-aware - Group by implementation phase
5. Learning path - Suggest next steps for deeper study
6. Concept mapping - Connect code to underlying CS concepts

---

🛠️ Supported Languages & Frameworks

• JavaScript/TypeScript: React, Node.js, Express
• Python: Django, FastAPI, Flask
• Go: Standard library, Gin, Echo
• Rust: Standard library, Actix
• Java: Spring Boot
• And more - Extensible pattern system

---

🤝 Contributing

Contributions welcome! To extend AntiVibe:

1. Add patterns to reference/language-patterns.md
2. Add resources to reference/resource-curation.md
3. Customize the template in templates/deep-dive.md

---

📖 Documentation

• Setup Guide - Detailed installation
• Skill Format - Claude Code skills

---

⚠️ License

MIT License - Use it, learn from it, share it.

---

_{Built with 🔥 for developers who actually want to understand code.}