Skip to content

Documentation Index

Welcome to the Code Review GPT GitLab documentation center!

πŸ“š Project Description

Code Review GPT GitLab leverages large language models (including but not limited to GPT and DeepSeek) to help development teams automate code reviews in GitLab and improve engineering efficiency.

Key Highlights ✨

  • πŸ€– Intelligent Code Reviews – Automatically analyses code quality, flags potential issues, and suggests improvements.
  • πŸ”— Deep GitLab Integration – Purpose-built for GitLab; reviews are triggered automatically through webhooks.
  • 🎯 Multi‑model Support – Integrates multiple LLM providers (GPT, DeepSeek, etc.), including self-hosted options.
  • πŸ“’ Multi-channel Notifications – Sends review updates through DingTalk, Feishu, WeChat, and more.
  • βš™οΈ Highly Customisable – Tailor prompts, review rules, and workflow logic to match team standards.
  • πŸš€ Multi-Agent Architecture – Ongoing work enables collaborative agents to further raise review quality.
  • πŸ”’ Security Checks – Detects potential security issues to keep your codebase safe.

🎯 Main Capabilities

1. Automated Code Reviews

  • Automatically triggers when a GitLab merge request is created.
  • Analyses code changes and produces detailed reports with recommendations.
  • Supports multiple programming languages.

2. Flexible Model Configuration

  • Connect several LLM services at once.
  • Choose different models per project.
  • Supports self-hosted LLM deployments.

3. Smart Notification System

  • Delivers results to configured channels (DingTalk, Feishu, WeChat, etc.).
  • Lets you customise notification formats and content.
  • Provides near real-time progress updates.

4. Project Management

  • Manage multiple GitLab projects from one place.
  • Inspect historical review records and metrics.
  • Configure project-level settings with ease.

πŸ“– Additional Information

For more details, see: - Main project README: ../README.md - Chinese documentation: zh/