Codie — AI Code Review Assistant

Secure, efficient, and explainable AI platform for end-to-end code quality

---

✨ Key Features

🗺️ Repository-wide dependency mapping and hotspots
🔍 Dynamic runtime analysis for regressions
💬 Chat-first, explainable reviews with diffs and one-click fixes
🎯 Adaptive style learning (project-specific)
🧪 AI test generation (Java via Diffblue, JS/Python via GPT)
💰 ROI-ranked refactoring guidance
🛡️ CVE-enriched security remediation
🎛️ Confidence-scored noise filtering

This repository is a monorepo with backend, frontend, AI orchestration, infrastructure, and comprehensive documentation with automated changelogs and governance.

🏗️ Architecture Overview

graph TB
    subgraph "User Interface"
        UI["🎨 UI<br/>Monaco Editor + D3 Viz"]
    end
    
    subgraph "API Gateway"
        GW["🚪 API Gateway<br/>Authentication & Routing"]
    end
    
    subgraph "Core Services"
        AUTH["🔐 Security<br/>User Management"]
        ANALYSIS["🔍 Analysis<br/>Static & Runtime"]
        AI["🤖 AI Orchestrator<br/>LLM Integration"]
        REVIEW["📝 Review<br/>Code Analysis"]
    end
    
    subgraph "Testing & Quality"
        TEST["🧪 Tests<br/>Generation & Execution"]
        METRICS["📊 Metrics<br/>Quality Scoring"]
    end
    
    subgraph "Data Layer"
        DB[("🐘 PostgreSQL<br/>Application Data")]
        CACHE[("🔴 Redis<br/>Session & Cache")]
        STORAGE[("📦 MinIO<br/>File Storage")]
    end
    
    UI --> GW
    GW --> AUTH
    GW --> ANALYSIS
    GW --> AI
    GW --> REVIEW
    
    ANALYSIS --> TEST
    AI --> REVIEW
    REVIEW --> METRICS
    
    AUTH --> DB
    ANALYSIS --> DB
    AI --> CACHE
    REVIEW --> DB
    TEST --> STORAGE
    METRICS --> DB

Loading

Architecture diagram showing the core components and data flow

📁 Monorepo Structure

Codie/
├── 🔧 codie-backend/     # FastAPI microservices
│   ├── API Gateway
│   ├── Analysis orchestration
│   ├── Static/runtime analysis
│   ├── Style ML & LLM review
│   └── Security & scoring
├── 🎨 codie-frontend/     # Next.js/React with Tailwind
│   ├── Radix UI components
│   ├── Monaco editor
│   └── D3/VisX visualizations
├── 🤖 ai-agents/          # AI prompts & automation
│   ├── Prompts & pipelines
│   ├── .clinerules
│   └── Codemods & scripts
├── 🐳 infra/              # Infrastructure & DevOps
│   ├── Docker Compose (local)
│   ├── Kubernetes manifests
│   └── CI/CD configurations
└── 📚 docs/               # Comprehensive documentation
    ├── PRDs & architecture
    ├── Service documentation
    ├── DevOps runbooks
    └── ADRs & OpenAPI specs

🚀 Quick Start

Prerequisites

• 🐳 Docker & Docker Compose
• 🟢 Node.js 18+
• 🐍 Python 3.11+
• 💾 Git

Installation

1. Clone the repository

   git clone https://github.com/Cyril-36/Codie.git
   cd Codie

2. Set up environment

   # Copy environment templates
   cp .env.sample .env
   # Fill in your secrets and configuration

3. Start the application

   docker compose up --build

4. Access the services • 🌐 Frontend: http://localhost:5174
   • 📡 Backend API: http://localhost:8000/docs
   • 🗄️ MinIO: http://localhost:9001
   • 🐘 PostgreSQL: localhost:5432
   • 🔴 Redis: localhost:6379

🛠️ Tech Stack

Backend

• Framework: FastAPI
• Language: Python 3.11+
• Database: PostgreSQL
• Cache: Redis
• Storage: MinIO
• Container: Docker

Frontend

• Framework: Next.js/React
• Styling: Tailwind CSS
• Components: Radix UI
• Editor: Monaco
• Visualization: D3, VisX

AI & ML

• LLM Integration: GPT-4, Claude
• Code Analysis: AST parsing
• Test Generation: Diffblue (Java)
• Style Learning: Custom ML models

DevOps & Infrastructure

• Containerization: Docker, Docker Compose
• Orchestration: Kubernetes
• CI/CD: GitHub Actions
• Monitoring: Custom metrics

📸 Screenshots

🏠 Dashboard Overview

Coming soon - Dashboard interface

📊 Code Analysis Report

Coming soon - Analysis report view

💬 AI Chat Interface

Coming soon - AI chat interface

Screenshots will be updated with actual application images

🛡️ Security & Quality

Security Features

• 🔐 TLS encryption in production
• 🎫 Least-privilege tokens
• 📦 Sandboxed runtimes
• 🔍 SBOM & vulnerability scans
• 🤐 Secret/PII redaction to LLMs

Code Quality

• 🐍 Python: ruff, black, mypy
• 📜 TypeScript: ESLint, strict mode
• 🧪 Testing: Unit, integration, E2E
• 📈 Coverage gates enforced

Documentation Governance

• 📝 Keep a Changelog + SemVer
• 📚 Per-module CHANGELOGs
• 🤖 CI-enforced docs updates
• 🔍 OpenAPI sync with Spectral lint
• 📊 Optional metrics: Docs/Test Coverage, Performance, Security

🤝 Contributing

We welcome contributions! Please see our Contributing Guide for details.

Development Setup

Backend Development:

cd codie-backend
python3 -m pip install -r requirements.txt
# Configure optional probes via .env
export ENABLE_POSTGRES_PROBE=true
export POSTGRES_DSN=postgresql://postgres:postgres@localhost:5432/postgres
export ENABLE_REDIS_PROBE=true
export REDIS_URL=redis://localhost:6379/0

uvicorn app.main:app --reload

Health Checks:

curl http://localhost:8000/livez   # Liveness probe
curl http://localhost:8000/readyz  # Readiness probe

Frontend Development:

cd codie-frontend
npm install
npm run dev

📋 Contribution Guidelines

• 🔄 Fork the repository
• 🌟 Create a feature branch
• ✅ Add tests for new features
• 📝 Update documentation
• 🚀 Submit a pull request

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

📞 Contact & Support

💬 Get in Touch

Contact Email: cyrilchaitanya@gmail.com