Pullwise is an open-source, self-hosted AI code review platform that combines static analysis (SAST) with large language models (LLMs) to provide intelligent, automated code reviews.

Code reviews are essential for software quality, but they're:

• Time-consuming - Senior developers spend hours reviewing PRs
• Inconsistent - Different reviewers catch different issues
• Expensive - Enterprise tools cost thousands per month
• Vendor lock-in - Proprietary solutions trap your data

Pullwise Community Edition (MIT Licensed):

• Free forever - No credit card, no time limits
• Self-hosted - Your code never leaves your infrastructure
• AI-Powered - Multi-model LLM support (GPT-4, Claude, local models)
• SAST Integration - SonarQube, ESLint, Checkstyle, PMD, SpotBugs
• Auto-Fix - One-click apply suggestions
• 200+ Plugins - Community-driven extensions

# Clone the repository
git clone https://github.com/integralltech/pullwise-ai.git
cd pullwise-ai/frontend

# Download the docker-compose.yml
wget https://pullwise.ai/docker-compose.yml

# Start all services
docker-compose up -d

# Access Pullwise
open http://localhost:8080

That's it! Pullwise is now running with:

• PostgreSQL database
• Redis cache
• Ollama (local LLM)
• Pullwise CE

• Docker 20.10+ and Docker Compose 2.0+
• 2 GB RAM minimum (4 GB recommended)
• 10 GB disk space
• Linux, macOS, or Windows with WSL2

Pullwise follows the GitLab open-core model:

→ Compare all editions

Pullwise combines the best of both worlds:

1. Static Analysis (parallel execution):

   • SonarQube (bugs, vulnerabilities, code smells)
   • ESLint (JavaScript/TypeScript)
   • Checkstyle (Java)
   • PMD (anti-patterns)
   • SpotBugs (bug patterns)

2. AI Review (with full context):

   • SAST results as baseline
   • Code graph analysis
   • Historical PR data
   • Custom team instructions

3. Smart Consolidation:

   • Deduplicates similar issues
   • Prioritizes by severity
   • Formats actionable comments

• Cloud models: GPT-4, Claude Sonnet, Gemini Pro via OpenRouter
• Local models: Llama 3, Mistral, Gemma via Ollama
• Cost optimization: Auto-routes to cheapest model for task
• Fallback: Graceful degradation when models fail

200+ community plugins extending:

• Language linters (Rust, Go, Python, PHP)
• Framework-specific rules (Laravel, Django, Spring)
• Custom checks for your codebase
• → Plugin Marketplace

• One-click apply for AI suggestions
• Safe preview before applying
• Rollback support
• Batch operations

Join 10,000+ developers using Pullwise:

• 5,000+ GitHub Stars
• 10,000+ Docker Pulls
• 200+ Community Plugins
• 1,000+ Discord Members

→ Join Discord

• Getting Started Guide
• Installation Guide
• Configuration
• Plugin Development
• API Reference

docker-compose up -d

kubectl apply -f k8s/community-edition/

helm repo add pullwise https://charts.pullwise.ai
helm install pullwise pullwise/pullwise-ce

We welcome contributions! Please see our Contributing Guide.

Areas where we need help:

• Language integrations (Rust, Go, Python, PHP)
• Plugin development
• Documentation improvements
• Bug reports and testing
• → Good First Issues

Community Edition - MIT License

Free to use, modify, and distribute. Forever.

Built with inspiration from:

• GitLab - Open-core business model
• SonarQube - SAST foundation
• CodeRabbit - AI review patterns
• Open Source Community - Tools and libraries