PRception is a GitHub App that automates pull request reviews using AI-powered code analysis. It provides intelligent, context-aware feedback, suggests improvements, and can automatically approve PRs that meet defined quality standards — saving developers time and ensuring consistent code quality.

✅ Automated Code Review – PRception analyzes code changes using GPT-4 and posts suggestions directly in the PR
✅ Automatic Approval – PRception can automatically approve PRs if they meet quality criteria
✅ Context-Aware Feedback – Provides line-by-line feedback based on code context
✅ Diff Analysis – Analyzes PR diffs to detect issues and improvements
✅ Customizable Review Rules – Define custom rules and quality standards
✅ Chat with PRception – Allows developers to ask follow-up questions directly in the PR comments and receive AI-powered responses.
✅ Secure Authentication – Uses GitHub App with JWT-based authentication
✅ Scalable – Works seamlessly with large repositories and complex codebases
✅ Multi-Repo Support – Supports reviewing across multiple repositories in an organization

Ensure you have the following installed:

• 🏆 Go (1.22 or higher)
• 🐳 Docker (latest version recommended)
• 🔑 GitHub App credentials (App ID and Private Key)
• 🧠 OpenAI API key

Create a .env file in the project root with the following:

GITHUB_APP_ID=your_app_id
GITHUB_PRIVATE_KEY=your_base64_encoded_private_key
OPENAI_API_KEY=your_openai_api_key

git clone https://github.com/yourusername/prception.git
cd prception

go mod download

1. Go to GitHub → Settings → Developer Settings → GitHub Apps

2. Create a new app

3. Set the following permissions:

• ✅ PR read/write
• ✅ Contents read

4. Subscribe to Pull Request event

5. Generate and download the private key

6. Install the app in your repository

./scripts/start.sh

You can deploy PRception using Docker for easy scalability and portability:

docker build -t prception .

docker run -d -p 8080:8080 --env-file .env prception

1. A PR is opened → GitHub sends a webhook to PRception
2. PRception fetches the PR diff and changed files
3. Code is analyzed using GPT-4 to detect issues and improvements
4. Intelligent review comments are posted directly on the PR
5. If the code meets the defined standards, PRception automatically approves the PR

PRception now supports an interactive chat feature!

Mention @prception in a PR comment Reply to ask follow-up questions about the code review. PRception will respond with context-aware answers based on the diff and file content. Example:

@prception Why is this loop inefficient?