Surgeon precisely reviewing a circuit board, symbolizing AI's detailed code analysis

Alibaba Open-Code-Review

Alibaba has open-sourced its internal AI code review assistant, open-code-review, which has served thousands of developers and identified millions of code defects. This CLI tool reads Git diffs, uses an LLM agent with tool-use capabilities, and…

Alibaba's open-source AI-powered code review tool offers precise, line-level comments and a hybrid architecture for efficient defect identification.

Developer in a large server farm, holding a device, showing the scale and impact of the AI tool

TL;DR

Alibaba has open-sourced its internal AI code review assistant, open-code-review, which has served thousands of developers and identified millions of code defects. This CLI tool reads Git diffs, uses an LLM agent with tool-use capabilities, and generates structured, line-level review comments. It achieves higher Precision and F1 scores with significantly fewer tokens compared to general-purpose agents.

Key facts

  • Maker: Alibaba
  • What's new: Open-sourced AI-powered code review CLI tool
  • License: Apache-2.0
  • Context: Battle-tested at Alibaba's scale, served thousands of developers
  • Benchmark: Higher Precision and F1, ~1/9 tokens vs. general-purpose agents
Merchants in a Chinese market inspecting goods, reflecting Alibaba's origin and the tool's precision

What is it?

Open Code Review is an AI-powered command-line interface tool for reviewing code. It originated as Alibaba Group's internal AI code review assistant. It has been used by tens of thousands of developers and has found millions of code defects.

How does it work?

The tool reads Git diffs and sends changed files to a configurable LLM via an agent. This agent uses tools to read full file contents, search the codebase, and inspect other changed files for context. It generates structured review comments with line-level precision.

Why does it matter?

This tool offers a more stable and predictable code review process compared to general-purpose agents. It combines deterministic engineering for critical steps with an agent for dynamic decisions. This approach ensures precise file selection and fine-grained rule matching.

Who is it for?

Developers, engineering teams, and organizations seeking an efficient and precise AI-powered code review solution.

Try it

Install via npm, GitHub Release binary, or source, then configure an LLM endpoint or use delegation mode.

Frequently asked questions

What is Open Code Review?

Open Code Review is an AI-powered command-line interface tool for reviewing code. It was developed internally at Alibaba Group and has been open-sourced for the community. It helps identify code defects with high precision.

How does Open Code Review work?

It reads Git diffs and sends changed files to a configurable LLM agent. This agent uses various tools to understand the code context and generates structured, line-level review comments. It can also scan entire files for auditing unfamiliar codebases.

Is Open Code Review open source?

Yes, Open Code Review is an open-source project. It is licensed under Apache-2.0 and its code is available on GitHub.

What are the benefits of using Open Code Review?

It offers significantly higher Precision and F1 scores compared to general-purpose agents, while consuming fewer tokens and completing reviews faster. It also provides precise file selection, smart file bundling, and fine-grained rule matching for stable review quality.

What models does Open Code Review support?

It is compatible with OpenAI and Anthropic models. Users need to configure a model endpoint to get started, or they can use delegation mode with their existing coding agent's model.

Sources

  1. GitHub

Follow news like this automatically

RAI AI watches news, X, Telegram and Reddit for the topics you choose and sends a short AI summary to the app, email or Telegram. Try RAI AI free for 7 days