Repolode
Submit
Nikita-Filonov

Nikita-Filonov/ai-review

AI-powered code review tool that posts inline, summary and reply comments on pull and merge requests using LLMs like OpenAI, Claude or Ollama.

Open-source alternative to:CodeRabbit

View on GitHubWebsite
Stars
601
Forks
103
Open issues
9
Last commit
last month

What is it?

AI Review runs as a CI/CD step on pull or merge requests and asks a large language model to review the diff. It can post line-by-line comments, a high-level summary, and follow-up replies inside existing review threads. The tool works with GitLab, GitHub, Bitbucket Cloud, Bitbucket Server, Azure DevOps and Gitea, and can use OpenAI, Claude, Gemini, Ollama, Bedrock, OpenRouter or Azure OpenAI as the model backend.

Configuration is done through a .ai-review.yaml, .ai-review.json or environment variables, with CLI subcommands such as run-inline, run-summary and run-context for running individual review modes. An optional agent mode lets the model run shell commands like ls, cat, rg and git to explore the repository before writing its review. Prompts for inline, context and summary reviews can be customized to match a team's coding guidelines.

The tool is installed from PyPI as xai-review or pulled as a Docker image, and ships with ready-made CI templates for GitHub Actions and GitLab CI/CD. When Ollama is used, requests go to a local or self-hosted runtime, so the review can run without sending code to an external API.

Key features

  • 01Posts inline comments, summary reviews and AI replies in existing threads.
  • 02Supports GitHub, GitLab, Bitbucket Cloud and Server, Azure DevOps and Gitea.
  • 03Switches between OpenAI, Claude, Gemini, Ollama, Bedrock, OpenRouter and Azure OpenAI.
  • 04Agent mode lets the model explore the repository with shell commands before reviewing.
  • 05Customizable prompt templates for inline, context and summary reviews.
  • 06Configurable through YAML, JSON or environment variables for CI overrides.

Use cases

  • Adding automated first-pass review comments to GitLab or GitHub merge requests.
  • Running reviews against a self-hosted Ollama model so code stays inside your infrastructure.
  • Enforcing a team's coding guidelines through custom review prompts.
  • Triggering review modes manually from a GitHub Actions workflow.

Quick start

  1. 1

    Install via pip from PyPI

    pip install xai-review
  2. 2

    Or run directly via Docker

    docker run --rm -v $(pwd):/app nikitafilonov/ai-review:latest ai-review run-summary
  3. 3

    Create a .ai-review.yaml configuration file in your project root with your LLM and VCS settings

Things to know

  • Running reviews requires an API token for the chosen LLM provider, unless you run a local Ollama instance.
  • Reviews are generated by an external LLM, so comment quality depends on the model and configuration.

In collections

AI code review toolsAI coding assistants & agentsRun LLMs locallyTesting frameworks & test automation tools

Topics

#code-review#llm#ci-cd#gitlab#github#ollama#openai#developer-tools#ai#ai-code-review

FAQ

What is ai-review?
AI-powered code review tool that posts inline, summary and reply comments on pull and merge requests using LLMs like OpenAI, Claude or Ollama. AI Review runs as a CI/CD step on pull or merge requests and asks a large language model to review the diff. It can post line-by-line comments, a high-level summary, and follow-up replies inside existing review threads. The tool works with GitLab, GitHub, Bitbucket Cloud, Bitbucket Server, Azure DevOps and Gitea, and can use OpenAI, Claude, Gemini, Ollama, Bedrock, OpenRouter or Azure OpenAI as the model backend.
Is ai-review free to use? What license does it have?
ai-review is open source under the Apache-2.0 license. Check the license terms before using it in commercial products.
Is ai-review still maintained?
Updated within 6 months. The last commit to ai-review was on Oct 3, 2026.
What can I use ai-review for?
Adding automated first-pass review comments to GitLab or GitHub merge requests. Running reviews against a self-hosted Ollama model so code stays inside your infrastructure. Enforcing a team's coding guidelines through custom review prompts. Triggering review modes manually from a GitHub Actions workflow.
How do I install ai-review?
Install via pip from PyPI: pip install xai-review Or run directly via Docker: docker run --rm -v $(pwd):/app nikitafilonov/ai-review:latest ai-review run-summary Create a .ai-review.yaml configuration file in your project root with your LLM and VCS settings
Where can I get ai-review?
The source code and installation instructions are on GitHub: https://github.com/Nikita-Filonov/ai-review

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects