Description
CC Auto Review is a macOS application that enhances the GitHub pull request review process by leveraging your local Claude Code. It aims to reduce the manual effort involved in reviewing numerous small pull requests, allowing developers to focus their attention on more critical or complex changes.
Read More
The application provides AI findings directly beside the code diff, making it easy to understand the context of each suggestion. Every pull request is assigned a risk score ranging from 0 to 100. This score is calculated based on factors such as size, spread, sensitive paths, risky code patterns, and test coverage gaps. The baseline score is determined from the diff alone, ensuring consistency regardless of the PR description. Claude can then adjust this score by up to 20 points, citing specific file and line numbers, and can escalate serious problems to a score of at least 60.
Users have full control over the AI's suggestions. Findings are presented as drafts, which can be accepted, rejected, or edited before being posted to GitHub from the user's own account. The app also allows users to ask Claude for explanations or further insights into specific findings or code sections, incorporating these discussions into the draft comments. A crucial safety feature is that nothing is posted if the PR head has moved since the review was conducted.
An optional autopilot mode can automatically review and post comments for less risky pull requests, defined by criteria such as a risk score below 50, or specific paths that are excluded from AI review (e.g., CI, migrations, secrets). This feature runs only while the app is open and posts reviews as automated comments from the user's GitHub account. It's important for users to be aware of Anthropic's terms or use an API key when enabling autopilot, as it runs Claude Code automatically.
CC Auto Review operates by reading only the changed files and diffs in a temporary directory, without cloning the entire repository or requiring an API key setup within the app itself. It integrates with your existing `gh` (GitHub CLI) login and `claude` (Claude Code) installation. The app supports review profiles, allowing users to set perspectives, team guidelines, and finding limits, and can re-check earlier comments against current code. It also reads repository conventions from `CLAUDE.md`, `AGENTS.md`, and `REVIEW.md` from the base branch. The application is free and open source under the MIT license, with review costs tied to the user's Claude Code plan or API account.
- CategoryAI Code Review & Testing
- Pricing ModelFree
- DeploymentLocal
- PlatformsmacOS
CC Auto Review's Core Features
Reviews GitHub pull requests with local Claude Code
Displays AI findings directly beside the code diff
Assigns a 0-100 risk score to every pull request
Optional autopilot for automated review of low-risk PRs
Allows users to accept, reject, or edit AI findings before posting
Supports asking Claude for explanations or code insights
Prevents posting if PR head has moved since review
Reads only changed files in a temporary directory (no cloning)
Integrates with existing GitHub CLI and Claude Code installations
Supports custom review profiles and team guidelines
Re-checks prior review comments against current code
Reads repository conventions from CLAUDE.md, AGENTS.md, REVIEW.md
Monitors PRs for author pushes or replies, re-running reviews
Provides desktop notifications for review updates
Free and open source under MIT license
How to use CC Auto Review?
Install GitHub CLI: Get `gh` and sign in with `gh auth login`.
Install Claude Code: Obtain Claude Code 2.1.314 or later.
Download and Install App: Open the DMG and drag CC Auto Review to Applications (macOS 11+ required).
Open a Pull Request: The app will automatically score the PR and display AI findings.
Review Findings: Accept, reject, or edit Claude's suggestions.
Post Comments: Publish your approved findings to GitHub from your account.
Enable Autopilot (Optional): Turn on autopilot for automated review of less risky PRs, ensuring the app window remains open.
CC Auto Review's Use Cases
- Automated PR Risk Scoring
- AI-Assisted Code Review
- Focus on High-Risk PRs
- Customizable Review Guidelines
- Ensuring Code Quality
- Continuous Code Monitoring
- Developer Productivity
- Consistent Review Feedback
