What is OpenCodeReview?
OpenCodeReview is an AI-powered code review CLI that originated as an internal review assistant at Alibaba Group. It reads Git diffs and sends changed files to a configurable language model through an agent that can read complete files, search the repository, and inspect related changes. Full-file scanning also supports reviews of unfamiliar codebases or directories without a meaningful diff.
Its hybrid architecture separates review orchestration from model reasoning. Deterministic logic selects files, groups related files into isolated review units, and matches rules to file characteristics. Separate modules handle comment positioning and reflection, while the agent retrieves context and evaluates potential defects. Built-in rules address issues including null pointer exceptions, thread safety, cross-site scripting, and SQL injection.
The tool can be installed through npm, installation scripts, release binaries, or a source build. Standard execution requires model configuration; delegation mode instead uses a host coding agent's model while OpenCodeReview handles file selection and rule resolution. Reviews can export JSON results, resume interrupted sessions, and be inspected through a browser-based session viewer.
OpenCodeReview screenshots
Explore the interface. Open a screenshot to view it at full size.
Source: official OpenCodeReview project page.
Key capabilities
- Diff review Analyzes workspace changes, branch ranges, and individual commits with structured comments tied to code locations.
- Full-file scanning Reviews complete files, selected directories, or an entire repository without requiring Git history.
- Repository context Reads full files, searches the codebase, and examines related changes during review.
- Rule orchestration Matches review rules to files and groups related files into isolated review units that support concurrent execution.
- Agent delegation Uses a configured model directly or delegates review reasoning to a compatible host coding agent.
- Session handling Resumes interrupted reviews, exports JSON results, and provides browser-based replay with fixed or ignored comment states.
Where OpenCodeReview fits
OpenCodeReview fits developer-side reviews before commits, branch-change checks, and full-file audits of unfamiliar repositories. Its documented CI/CD integrations include GitHub Actions, GitLab CI, GitFlic CI, and Gerrit. Coding-agent integrations allow reviews to run within an existing assistant workflow rather than only as a separate terminal command.
Before deployment, teams should verify Git prerequisites, file filters, rule coverage, model connectivity, and the handling of repository contents sent to the selected model endpoint. Standard mode requires model credentials and configuration, while delegation mode relies on the host agent's model. Findings should be assessed alongside existing review and testing practices because the project's stated preference for precision over recall can leave defects unreported.
Install
npm install -g @alibaba-group/open-code-reviewReviews of OpenCodeReview
Be the first to share your experience with OpenCodeReview.
Write a review
Loading review form…
Alternatives to OpenCodeReview
AiderOpen-source command-line tool for editing code with LLMs inside a local git repository. It builds a repository map, applies model edits to files, commits each change, runs linters and tests, and works with most hosted and local models.AAirshipAirship is a local CLI with a browser visual editor for existing web apps; Claude Code, Codex or OpenCode implement the selected changes.
ClineApache-2.0 coding agent from Cline Bot Inc. that edits files, runs commands and uses MCP tools with per-action approval. Ships as a VS Code extension, JetBrains plugin, npm CLI, desktop app and TypeScript SDK, with BYOK or Cline-billed models.