Open Code Review review

Q: How do I install Open Code Review?

Run npm install -g @alibaba-group/open-code-review. It needs Node.js >=14.

Q: How do I review a branch with Open Code Review?

Use ocr review --from main --to feature-branch. This reviews the feature branch's changes since it diverged from main. You can also review a single commit with ocr review --commit abc123.

Q: Can Open Code Review resume an interrupted review?

Yes. Run ocr session list to find the session. Then pass --resume to ocr review or ocr scan.

Open Code Review is an AI code review CLI that combines deterministic pipelines with an LLM agent to produce line-level review comments on Git changes.

Updated Nov 1, 2026. Review written Oct 22, 2026 from the project's own sources; usage data refreshed every Monday.

TL;DR

Open Code Review is a code review CLI that started as Alibaba Group's internal AI code review assistant. It suits teams that want structured, line-level review comments from an LLM they configure. It uses deterministic engineering for file selection, bundling and rule matching, and an agent for dynamic context retrieval. The README describes this as favoring precision over recall.

49.2kGitHub stars
+1knew stars, week of Oct 25
~167.8kdownloads a week, −0% vs 4-week avg
#57in Coding agents this week

Open Code Review at a glance

Category: Coding agents
Maintained by: Alibaba (organization, @alibaba on GitHub)
License: Apache 2.0
Language: Go, JavaScript, TypeScript
Repository: github.com/alibaba/open-code-review
Website: open-codereview.ai
npm package: @alibaba-group/open-code-review 1.13.2
Latest release: v1.13.2 on GitHub, Oct 20, 2026
First published: May 18, 2026

What Open Code Review does

Open Code Review reads Git diffs and generates structured review comments with line-level precision using a configurable LLM.
Open Code Review's agent can read full files, search the codebase and inspect other changed files for context.
Open Code Review reviews entire files with `ocr scan` for auditing unfamiliar codebases or directories without a meaningful diff.
Open Code Review includes a built-in multi-language ruleset covering NPE, thread-safety, XSS and SQL injection.
Open Code Review bundles related files into sub-agents with isolated context, which supports concurrent review of large changesets.
Open Code Review offers a delegation mode where your AI coding agent performs the review using its own LLM.
Open Code Review provides a Session Viewer to browse and replay review sessions and mark comments as fixed or ignored.

How it works

Open Code Review runs as the `ocr` command in a Git repository. It connects to an OpenAI- or Anthropic-compatible model endpoint that you configure. It also integrates with Claude Code, Codex, Cursor, Kimi Code and OpenCode through plugins. It runs in CI through GitHub Actions, GitLab CI, GitFlic CI and Gerrit. It can be extended with MCP servers and supports OpenTelemetry.

How to install and start using Open Code Review

With npm

npm install -g @alibaba-group/open-code-review

Requirements

Git 2.41 or newer
Node.js 14 or newer for the npm package
A configured LLM provider and API key, unless you use delegation mode

First steps

Run `ocr config provider` to select a built-in provider or add a custom one, including API key entry.
Run `ocr config model` to pick a model for the active provider. The setup then tests connectivity automatically.
Run `ocr review` in your project to review all staged, unstaged and untracked changes.
Run `ocr review --format json --output result.json` to save results to a file, which the README recommends for AI host agents.

Who Open Code Review is for

Teams wanting automated pull request review in CI with GitHub Actions, GitLab CI or Gerrit
Developers using Claude Code, Codex, Cursor or OpenCode who want review commands inside their agent
Reviewers who prefer fewer false alarms over catching every possible issue
Developers auditing unfamiliar codebases with full-file scans

Limitations to know

Open Code Review has lower recall than general-purpose agents in its own benchmark, a deliberate trade-off favoring precision.
Open Code Review requires an LLM configuration before reviewing, unless you use delegation mode.
Open Code Review depends on Git 2.41 or newer for diffs, code search and repository operations.

Is Open Code Review growing? Usage data

In the week of Oct 25–31, 2026, Open Code Review ranked #156 on the Devtoolwatch Board and #57 in Coding agents, with a momentum score of 51 out of 100. Its main package had about 167.8k downloads that week, −0% against its own 4-week average.

New GitHub stars per week for the last 12 weeks, from GitHub's stargazer history (today's stargazers by the week they starred).

Show the weekly numbers

New GitHub stars per week for Open Code Review
Week of	New stars
Aug 9, 2026	841
Aug 16, 2026	637
Aug 23, 2026	487
Aug 30, 2026	348
Sep 6, 2026	844
Sep 13, 2026	15k
Sep 20, 2026	3.8k
Sep 27, 2026	2k
Oct 4, 2026	2.4k
Oct 11, 2026	1.7k
Oct 18, 2026	733
Oct 25, 2026	1k

On Hacker News

Open Code Review – An AI-powered code review CLI tool, 284 points and 73 comments (Jun 5, 2026)

The Board ranks tools by four signals each week: new stars, new stars for the repo's size, added downloads and download growth beyond the market. How the ranking works.

Open Code Review alternatives

Other AI coding agents on the Board in the week of Oct 25–31, 2026, by momentum:

Coding agents ranked by momentum, the week of Oct 25–31, 2026
Rank	Tool	Stars	Downloads/wk
#1	CindyCindy is an open-source AI agent app for desktop and mobile that combines Claude Code, Codex and Pi across your connected devices.	8.8k	–
#2	OpenMausBotOpenMausBot is an open-source desktop chat app where each contact is a local Claude, Codex, or Grok CLI agent with its own computer.	9.6k	–
#3	OrcaOrca is a desktop agent development environment for running Claude Code, Codex, and other CLI coding agents in parallel, each in its own…	111.7k	–
#4	OpenBotOpenBot is an open-source, self-hosted template for AI coworkers that each get their own browser, files and granted tools, with every…	9.1k	–
#5	Agent ReachAgent Reach is a Python CLI that lets AI agents search and read platforms such as Twitter, Reddit, YouTube, GitHub, Bilibili and…	107.2k	–
#6	Skill RecorderSkill Recorder is a desktop app that records your on-screen work session and turns it into a reusable agent Skill or Automation.	5.6k	–

All AI coding agents, ranked

Recent Open Code Review releases

v1.13.2, Oct 20, 2026
v1.13.1, Oct 17, 2026
v1.13.0, Oct 13, 2026
v1.12.13, Oct 8, 2026
v1.12.12, Oct 5, 2026

Open Code Review FAQ

Is Open Code Review free and open source?

Yes. Open Code Review's source code is on GitHub (alibaba/open-code-review) under the Apache 2.0 license, so you can use it for free.

How do I install Open Code Review?

Run npm install -g @alibaba-group/open-code-review. It needs Node.js >=14.

Is Open Code Review still maintained?

Yes. The alibaba/open-code-review repository was last updated on Oct 20, 2026, and the latest release, v1.13.2 on GitHub, came out on Oct 20, 2026.

How popular is Open Code Review?

Open Code Review has 49.2k GitHub stars as of Nov 1, 2026. It gained 1k new stars in the week of Oct 25–31, 2026, against a 12-week average of 2.5k a week. Its main package had about 167.8k downloads that week, −0% against its own 4-week average. That made it #57 among AI coding agents on the Devtoolwatch Board.

What are the best alternatives to Open Code Review?

Other AI coding agents on the Devtoolwatch Board in the week of Oct 25–31, 2026, by momentum: Cindy (#1), OpenMausBot (#2), Orca (#3), OpenBot (#4), Agent Reach (#5). The alternatives table above has their numbers.

Which LLMs does Open Code Review work with?

Open Code Review works with OpenAI- and Anthropic-compatible endpoints. You can select a built-in provider or add a custom one with ocr config provider. Release notes also mention OpenRouter model overrides and ChatGPT plan reviews.

Can Open Code Review run without its own API key?

Yes. In delegation mode, your coding agent runs the review using its own LLM. Open Code Review handles file selection and rule resolution, and no OCR API key is required.

How do I review a branch with Open Code Review?

Use ocr review --from main --to feature-branch. This reviews the feature branch's changes since it diverged from main. You can also review a single commit with ocr review --commit abc123.

Can Open Code Review resume an interrupted review?

Yes. Run ocr session list to find the session. Then pass --resume <session-id> to ocr review or ocr scan.

How is Open Code Review different from using a general-purpose agent for review?

Open Code Review uses deterministic logic for file selection, bundling, rule matching and comment positioning. It uses the agent only for dynamic decisions and context retrieval. The README says this design addresses incomplete coverage, position drift and unstable quality.

Sources

Open Code Review README badge

Maintainers can show the tool's weekly Board rank in their README:

[![Devtoolwatch](https://index.devtoolwatch.com/badge/alibaba-open-code-review.svg)](https://index.devtoolwatch.com/tools/alibaba-open-code-review/)

How this page is made: Devtoolwatch writes each review from the project's README, releases, package metadata and website, with an AI model, then checks every claim against the source it cites and drops what the sources don't support. The numbers come from Devtoolwatch's daily GitHub snapshots, GitHub's stargazer history, and the npm and PyPI download APIs. Spotted an error? Email [email protected].