Skip to content

Folders and files

NameName
Last commit message
Last commit date

Latest commit

Β 

History

4 Commits
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 

Repository files navigation

πŸ” AI Code Reviewer

An AI-powered code review tool built with Claude API. Paste your code, get instant structured feedback on bugs, security issues, and improvements β€” in seconds.

πŸš€ What It Does

Developers spend hours on manual code reviews. This tool uses Claude AI to instantly analyse your code and return structured feedback β€” so you can focus on building, not reviewing.

You paste code β†’ Claude analyses it β†’ You get a structured review in under 5 seconds.


✨ Features

  • Bug Detection β€” Identifies logic errors, null pointer risks, and common pitfalls
  • Security Analysis β€” Flags injection vulnerabilities, exposed secrets, and unsafe patterns
  • Code Quality β€” Suggests improvements for readability, naming, and structure
  • Best Practices β€” Checks against language-specific conventions
  • Multi-language Support β€” Works with JavaScript, Java, Python, TypeScript, and more

πŸ—οΈ How It Works

  1. User pastes code into the web interface
  2. Frontend sends code to Express backend via POST /api/review
  3. Backend constructs a structured prompt and sends to Claude API
  4. Claude analyses the code and returns categorised feedback
  5. Frontend renders issues organised by category

πŸ› οΈ Tech Stack

Layer Technology
Backend Node.js + Express
AI Anthropic Claude API
Frontend HTML, CSS, Vanilla JS
API Style REST

βš™οΈ Getting Started

Prerequisites

Installation

Clone the repository and install dependencies:

git clone https://github.com/manasa-shivananda/ai-code-reviewer.git
cd ai-code-reviewer
npm install

Configuration

Create a .env file and add your API key:

ANTHROPIC_API_KEY=your_api_key_here
PORT=3000

Run

node src/index.js

Open http://localhost:3000 in your browser.


πŸ“Έ Demo

AI Code Reviewer


🧠 What I Learned

  • Claude API integration β€” prompt engineering for structured, consistent output
  • System prompt design β€” crafting prompts that return reliable, parseable responses
  • REST API design β€” clean separation between AI logic and server layer
  • Error handling for AI responses β€” managing token limits, timeouts, and malformed outputs

πŸ—ΊοΈ Roadmap

  • Structured JSON output with line-level issue mapping
  • Support for file upload (.js, .java, .py)
  • Side-by-side diff view with suggested fixes
  • GitHub PR integration via webhook

πŸ‘©β€πŸ’» About

Built by Manasa Shivananda β€” Full-Stack Developer specialising in AI-powered tooling.

AI Portfolio Series:

  • βœ… Project 1: AI Code Reviewer (this project)
  • πŸ”„ Project 2: AI Document Q&A Tool β€” in progress

πŸ“„ License

MIT License

About

πŸ” AI-powered code reviewer that catches bugs, security vulnerabilities and performance issues using Claude API

Resources

Stars

Watchers

Forks

Releases

Packages

Contributors

Languages