alibaba/open-code-review

What it solves

Open Code Review is an AI-powered CLI tool designed to automate code reviews. It addresses common failures of general-purpose AI agents, such as incomplete coverage of large changesets, inaccurate line-number positioning (position drift), and inconsistent review quality caused by purely language-driven prompts.

How it works

The tool uses a hybrid architecture that combines deterministic engineering with an AI agent:

• Deterministic Engineering: Handles critical constraints like precise file selection, smart bundling of related files into isolated sub-agents for concurrent review, and template-based rule matching to focus the model's attention.
• AI Agent: Manages dynamic decision-making and context retrieval using scenario-tuned prompts and a specialized toolset distilled from production data.
• Execution: It reads Git diffs, sends changed files to a configurable LLM via the agent, and generates structured, line-precise review comments. It also supports a scan mode for reviewing entire files without a diff.

Who it’s for

It is intended for developers and teams who want to integrate AI-driven code reviews into their local workflow or CI/CD pipelines (GitHub Actions, GitLab CI, etc.) and those using AI coding agents like Claude Code, Cursor, or Kimi Code.

Highlights

• High Precision: Optimized to reduce false alarms compared to general-purpose agents.
• Token Efficiency: Consumes significantly fewer tokens (approximately 1/9th) than general-purpose agents for similar tasks.
• Line-Level Precision: Uses independent positioning and reflection modules to ensure comments match the actual code location.
• Flexible Deployment: Supports various LLM providers and a "Delegation Mode" where the host AI coding agent performs the review using its own LLM.
• Broad Integration: Compatible with multiple AI coding agents and CI/CD platforms.