Skip to content

KhomenkovDev/khomdev_code_reviewer

Repository files navigation

KhomDev Code Reviewer

KhomDev Code Reviewer Icon

Screenshot 1 Screenshot 2 An AI-powered local code reviewer with a PyQt6 Graphical User Interface. This tool allows you to scan a directory, analyze your code using LLMs, and review feedback cleanly in a desktop application.

Installation

You can install ai-code-reviewer directly using pip:

pip install .

For development, install in editable mode:

pip install -e .

Usage

Once installed, you can start the application by running the ai-reviewer command from your terminal:

ai-reviewer

Running the Web App

You can also run the modern, glassmorphism web application version locally:

  1. Add your GEMINI_API_KEY to the .env file.
  2. Run the web server via Uvicorn:
uvicorn web.app:app --reload
  1. Open http://localhost:8000 in your browser.

Deployment & Packaging

1. Desktop Packaging

To package the native desktop application for your operating system (Windows/macOS/Linux):

  1. Install PyInstaller: pip install pyinstaller
  2. Run the build script: pyinstaller --onefile --windowed AI\ Code\ Reviewer.spec
  3. Locate your compiled executable in the /dist directory.
  4. Upload this file to your GitHub Repository under Releases.

2. Google Cloud Deployment (Web App)

The web application is fully prepared for containerized deployment to Google Cloud Run.

  1. Build and submit your Docker image using Google Cloud Build:
gcloud builds submit --tag gcr.io/YOUR_PROJECT_ID/khomdev-code-reviewer
  1. Deploy the image to Cloud Run:
gcloud run deploy khomdev-code-reviewer \
  --image gcr.io/YOUR_PROJECT_ID/khomdev-code-reviewer \
  --platform managed \
  --allow-unauthenticated \
  --set-env-vars="GEMINI_API_KEY=your_actual_key_here"

Alternatively, you can run the run.py script from the repository root to launch the desktop app locally:

python run.py

Features

  • Local directory scanning to pick up code files
  • AI analysis utilizing the google-genai library
  • PyQt6-based clean graphical user interface
  • Rendering of markdown-formatted responses from the LLM

About

A high-performance PyQt6 desktop application integrating local LLM capabilities to automate code review, enforce standards, and identify security vulnerabilities locally.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors