https://github.com/icelaterdc/ai-code-review-pro
Extensible, AI-powered CLI and GitHub Action with plugin support, multi-language review & rich outputs.
https://github.com/icelaterdc/ai-code-review-pro
ai-code-review ai-powered ai-powered-tools ai-review automated-code-review cicd code-quality extensible github-actions gpt-4 jest-test jest-tests lint multi-language-code-snippets openai-cli plugin-based review-climate terminal-code winston-logger winston-logging
Last synced: 1 day ago
JSON representation
Extensible, AI-powered CLI and GitHub Action with plugin support, multi-language review & rich outputs.
- Host: GitHub
- URL: https://github.com/icelaterdc/ai-code-review-pro
- Owner: icelaterdc
- Created: 2025-06-25T00:29:39.000Z (6 days ago)
- Default Branch: main
- Last Pushed: 2025-06-25T00:38:49.000Z (6 days ago)
- Last Synced: 2025-06-25T01:36:52.471Z (6 days ago)
- Topics: ai-code-review, ai-powered, ai-powered-tools, ai-review, automated-code-review, cicd, code-quality, extensible, github-actions, gpt-4, jest-test, jest-tests, lint, multi-language-code-snippets, openai-cli, plugin-based, review-climate, terminal-code, winston-logger, winston-logging
- Language: JavaScript
- Homepage:
- Size: 15.6 KB
- Stars: 2
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# AI Code Review Pro
AI Code Review Pro is a powerful, extensible CLI tool and GitHub Action that automates code review using OpenAI’s language models. It helps teams maintain code quality, enforce style guidelines, and catch potential bugs before merging.
---
## 🛠️ Features
* **Modular Architecture**: Core engine, CLI layer, plugin system and output handlers are decoupled.
* **Multi-Language Support**: JavaScript, TypeScript, Python, Go, Rust, Java (easily extendable).
* **Plugin System**: Write and share custom linting and style rules.
* **Flexible Outputs**: Terminal, JSON, HTML, Markdown report formats.
* **Configurable Thresholds**: Control sensitivity for style, security and best-practice suggestions.
* **GitHub Action**: Automatically review pull requests on `opened`, `synchronize`, or `reopened` events.
* **Detailed Logs & Metrics**: Structured logging via Winston, metrics for review duration and issue counts.
* **Built-In Testing**: Jest-based test suite to validate parser and plugin behavior.
---
## 📋 Prerequisites
* **Node.js** v16 or later
* **npm** v8 or later
* An **OpenAI API Key** (GPT-4 capable)
* GitHub repository with `main` branch
---
## 🔧 Installation
1. **Clone the repository**
```bash
git clone https://github.com/icelaterdc/AI-Code-Review-Pro.git
cd AI-Code-Review-Pro
```
2. **Install dependencies**
```bash
npm install
```
3. **Copy environment template**
```bash
cp .env.example .env
```
4. **Set your OpenAI key** in `.env`
```text
OPENAI_API_KEY=sk-...
```
---
## ⚡ Quick Start
1. **Initialize default configuration**
```bash
npx AI-Code-Review-Pro init
```
2. **Run a review** on your source folder
```bash
npx AI-Code-Review-Pro review src/ --output html,json
```
3. **View report**
* Terminal output summarized
* `review-report.html` autogenerated in project root
---
## 💻 CLI Usage
```bash
Usage: ai-review-pro [options] [command]
Commands:
review Review code at specified file or directory
init Generate default `aireview.config.js`
plugin:install
Install a plugin from npm or local path
help [command] Display help for command
Options:
-V, --version output the version number
-h, --help display help for command
```
### Examples
* **Review single file**
```bash
npx AI-Code-Review-Pro review src/index.js
```
* **Review with JSON output**
```bash
npx AI-Code-Review-Pro review src/ -o json
```
* **Install community plugin**
```bash
npx AI-Code-Review-Pro plugin:install @my-org/review-plugin-security
```
---
## ⚙️ Configuration
Default config file: `aireview.config.js` in project root.
```js
module.exports = {
openaiModel: 'gpt-4o-mini', // Model name
threshold: 0.75, // Confidence threshold (0.0 - 1.0)
languages: ['js','ts','py'], // Extensions to review
output: ['terminal','html'], // Output formats
plugins: [ // Plugins to load
'./plugins/sample-plugin.js'
],
rules: { // Built-in rule overrides
'no-todo-comments': true,
'max-line-length': 120
}
};
```
* **openaiModel**: Change to any supported OpenAI model.
* **threshold**: Lower thresholds generate more suggestions.
* **languages**: File extensions to include.
* **output**: `terminal`, `json`, `html`, `markdown`.
---
## 🧩 Plugin Development
Create a plugin by extending the base `Plugin` class:
```js
import { Plugin } from 'ai-review-pro/src/plugins/base.js';
export default class MyCustomPlugin extends Plugin {
apply(review) {
const { text, issues } = review;
if (/eval\(/.test(text)) {
review.addIssue({
message: 'Avoid using eval(), it may introduce security risks.',
severity: 'error',
});
}
}
}
```
* Place plugin file under `plugins/`
* Reference in `plugins` array of config
* Plugins can manipulate `review.issues` and metadata
---
## 📊 Outputs & Reports
* **Terminal**: Colorized summary
* **JSON**: Raw data for integrations
* **HTML**: Styled report with sections per file
* **Markdown**: For GitHub PR comments
All reports are saved in `./review-report.[ext]` by default.
---
## 📈 Logging & Metrics
Using Winston for structured logs:
```js
{
level: 'info',
transports: [new winston.transports.Console()]
}
```
Metrics collected:
* Review start/end timestamps
* Files processed count
* Total issues found
---
## 🧪 Testing
Run unit tests:
```bash
npm test
```
Includes parser and plugin behavior tests. Coverage reports generated under `coverage/`.
---
## 🤗 Contributing
1. Fork the repo
2. Create feature branch: `git checkout -b feature/my-feature`
3. Commit changes: `git commit -m 'Add new plugin support'`
4. Push: `git push origin feature/my-feature`
5. Open a Pull Request
Please follow the [Code of Conduct](CODE_OF_CONDUCT.md) and [Contributor Guidelines](CONTRIBUTING.md).
---
## 📜 License
MIT © 2025 IceLater
---
## ❓ FAQ
**Q:** How do I exclude files or directories?
**A:** Add a `.aireviewignore` file with glob patterns to exclude.
**Q:** Can I review only staged changes?
**A:** Yes, use `--staged` flag in `review` command (coming soon).
**Q:** Which OpenAI models are supported?
**A:** Any model available through your API key; default is `gpt-4o-mini`.