open-code-review: an AI-powered code review CLI that combines deterministic engineering with LLM agents for high-precision feedback

open-code-review: an AI-powered code review CLI that combines deterministic engineering with LLM agents for high-precision feedback

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, "position drift" (where AI comments don't match the actual line of code), 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 hard constraints like precise file selection, smart bundling of related files to manage context, and template-based rule matching to keep the model focused.
  • AI Agent: Manages dynamic decision-making and context retrieval using scenario-tuned prompts and a specialized toolset distilled from production data.
  • Process: It reads Git diffs, sends changed files to a configurable LLM via an agent with tool-use capabilities, and generates structured, line-precise review comments. It can also perform full-file scans for auditing codebases without diffs.

Who it’s for

  • Developers who want high-precision AI code reviews in their local workflow.
  • DevOps/CI/CD Engineers looking to integrate automated AI reviews into GitHub Actions or GitLab CI pipelines.
  • Users of AI coding agents (like Claude Code, Cursor, or Codex) who want to add a specialized code review skill to their existing agent workflow.

Highlights

  • High Precision: Optimized to favor precision over noise, reducing false alarms compared to general-purpose agents.
  • Agent Integration: Seamlessly integrates as a plugin or slash command for Claude Code, Cursor, and Codex.
  • Flexible LLM Support: Supports various built-in providers and custom LLM endpoints (OpenAI and Anthropic protocols).
  • Line-Level Precision: Uses independent positioning and reflection modules to ensure comments are accurately placed.
  • CI/CD Ready: Provides JSON output for easy integration into automated pipelines.

Sources