Code Review GPT

Code Review GPT uses Large Language Models to review code in your CI/CD pipeline. It helps streamline the code review process by providing feedback on code that may have issues or areas for improvement.

It should pick up on common issues such as:

Dead code
Exposed secrets
Slow or inefficient code
Unreadable code

It can also be run locally in your command line to review staged files.

Just so you know, this is in alpha and should be used for fun only. It may provide helpful suggestions or they may be completely wrong.

Demo

code-review-gpt-3.mp4

Prerequisites

Node.js
Git
Github CLI (optional for setup tool)

Easy Setup (Github Actions)

In the root of your git repository run:

npm install code-review-gpt
npx code-review-gpt configure

Template CI Usage (GitHub Actions)

- name: Install code-review-gpt
   run: npm install code-review-gpt

- name: Run code review script
   run: npx code-review-gpt review --ci
   env:
      GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
      BASE_SHA: ${{ github.event.pull_request.base.sha }}
      GITHUB_SHA: ${{ github.sha }}
      OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}

See templates/pr.yml for an example.

Getting Started

Clone the repository:

git clone https://github.com/mattzcarey/code-review-gpt.git
cd code-review-gpt

Install dependencies:
```
npm install
```
Set up the API key:
- Rename the .env.example file to .env.
- Open the .env file and replace YOUR_API_KEY with your actual OPENAI API key.

When used globally you should run export OPENAI_API_KEY=YOUR_API_KEY (or similar for your operating system) in your terminal to set the API key.

Run the application:
```
npm start
```
Install the application globally (optional):
```
npm run build
npm install -g
```
This will allow you to run the application from anywhere on your machine.

Usage

Not installed globally

Run npm i code-review-gpt && npx code-review-gpt in the root directory of a git repository.

Installed globally

Run code-review-gpt in the root directory of a git repository.

Commands

code-review-gpt review - Runs the code review on the staged files.
code-review-gpt configure - Runs a setup tool to configure the application.

Options

--ci - Runs the application in CI mode. This will use the BASE_SHA and GITHUB_SHA environment variables to determine which files to review. It will also use the GITHUB_TOKEN environment variable to create a comment on the pull request with the review results.
--model - The model to use for the review. Defaults to gpt-4. You can use any openai model you have access to.

Roadmap

Make a more clever way to find the exact code to review
VSCode extension
Use some embeddings and vector store to build a knowledge graph of the repo to make better suggestions
Prompt engineering to refine the prompt
Build a prompt analysis tool
Support different LLMs... Private, HuggingFace, Azure etc.
Build out the cloud offering

sasigit7/code-review-gpt