How hewking/aiplayground works
Key Components to Include:
Trigger Layer
• GitLab CI/CD Pipeline
• Webhook Server
Core Application Layer
• Main Reviewer Component (gitlab_ai_reviewer.py)
• Configuration Management (src/config)
• Core Utilities (src/core)
Model Layer
• OpenAI Integration
• DeepSeek Integration
• Gemini Integration
GitLab Integration Layer
• Code Review Handler
• Webhook Handler
Components as intermediaries between the trigger layer and core application.
┌─────────────────── Triggers ───────────────────┐
• [GitLab CI/CD] ──────── [Webhook Server] │
• ┼──────────────────────┼─────────────┘
┌──────────┼──────────────────────┼─────────────┐
• ▼ ▼ │
• [GitLab Integration Layer] │
• Code Reviewer │
• Webhook Handler │
• ┬───────────────────────────┘
┌───────────────────┼───────────────────────────┐
• ▼ │
• [Core Application Layer] │
• ┌─────────────────────────┐ │
• GitLab AI Reviewer │ │
• Review Logic │ │
• File Filtering │ │
• Prompt Generation │ │
• ┘ │
• ┼───────────────────────────┘
┌───────────────────┼───────────────────────────┐
• ▼ │
• [AI Models Layer] │
• ┌─────────┬──────────┬─────────┐ │
• OpenAI │ DeepSeek │ Gemini │ │
• ┴──────────┴─────────┘ │
• ┘
Important Elements to Highlight:
Data Flow
• Flow of code review requests and responses
Component Relationships
• Inheritance relationships between AI models
• Configuration dependencies
• Async boundaries