Documentation Specialist
Three-tiered analysis finds typos, grammar issues, and missing documentation using dictionary lookups, pattern matching, and LLM intelligence
AI-Powered Code Review for Gitea, GitHub, and GitLab
🧙♂️ v1.1.0 - Interactive Setup Wizard
Automated documentation quality checks, inline PR reviews, and intelligent issue creation—powered by your local LLM or AWS Bedrock (Claude 4.5 with ZDR)
Everything You Need for Automated Code Review
Automated PR Reviews
Inline comments on pull requests with actionable feedback. Works with Gitea, GitHub, and GitLab platforms
Performance Optimized
Intelligent caching, circuit breakers, and parallel analysis ensure fast reviews even on large repositories
Flexible LLM Backends
Local options (Ollama, LM Studio, llama.cpp, vLLM) for complete control, or AWS Bedrock for enterprise Claude 4.5 with Zero Data Retention
Set It and Forget It
Connect drep to your Git platform with webhooks for fully automated code reviews. No manual intervention required.
Pre-commit → drep checks locally
✓ Checks staged files before commit
✓ Catches issues before push
Push to repo → drep scans code
✓ Detects documentation issues
✓ Creates GitHub/Gitea issues
✓ Tracks findings in database
Open pull request → drep reviews
✓ Analyzes code changes
✓ Posts inline comments
✓ Provides actionable feedback
Three Layers of Automation
- Pre-commit hooks - Catch issues locally before commit
- Webhook-driven scans - Automatic checks on every push
- PR reviews - Inline comments on pull requests
- Self-hosted - Your code stays on your infrastructure
Get Started in Seconds
bash
# via pip
pip install drep-ai
# via Homebrew (macOS/Linux)
brew tap slb350/drep
brew install drep-ai
bash
# Initialize configuration
drep init
# Check code locally (pre-commit friendly)
drep check --staged
# Scan a repository (local or remote)
drep scan /path/to/repo
drep scan slb350/drep
# Review a pull request
drep review --pr 123 --repo slb350/drep