toriihq

toriihq / claude-review-action

Public

Reusable GitHub Action for AI-powered code review with Claude

ai anthropic claude code-review github-actions
26
3
100% credibility
Found Mar 12, 2026 at 24 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Shell
AI Summary

A GitHub Action that automates AI-powered code reviews for pull requests using Claude, with support for custom rules, re-reviews, and configurable authority levels.

How It Works

1
📰 Discover AI code helper

You hear about a friendly AI tool that automatically checks code changes in your shared project updates, saving time on reviews.

2
📂 Add to your project

Place a ready-made setup note in your project's hidden instructions folder so it runs whenever needed.

3
🔑 Connect the AI brain

Share a private password with the AI service, letting it wake up and start thinking about your code.

4
📝 Share project secrets

Whisper details about your project's rules, style, and must-follow guidelines to make the AI's advice perfect for you.

5
🏷️ Set review trigger

Create a simple tag or mention phrase that wakes the AI whenever you're ready for a check.

6
🔍 Request a review

On a new project update, add the tag or say the magic words, and the AI jumps in to scan everything.

✅ Receive smart feedback

In just minutes, a thorough review appears with tips, fixes, and approvals, making your code stronger and team happier.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 24 to 26 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is claude-review-action?

This reusable GitHub action delivers AI-powered code reviews using Claude on pull requests. Add a simple workflow YAML, your Anthropic API key as a secret, and trigger reviews via a "claude-review" label or "@claude" comments—Claude posts structured feedback directly as PR reviews. Built in pure bash on top of Anthropic's claude-code-action, it feeds diffs, PR context, and custom rules to Claude for security checks, style enforcement, and verdict summaries.

Why is it gaining traction?

It stands out with smart re-review reconciliation—Claude tracks prior findings, author replies, and new commits, marking issues as fixed or open. Configurable authority levels let it comment, request changes, or even approve small clean PRs, plus built-in cost tracking and diff truncation handling for large changes. Developers love the zero-build setup and seamless integration with existing Claude Code apps for context like CLAUDE.md files.

Who should use this?

PR-heavy teams enforcing critical rules, like backend devs mandating parameterized queries or auth checks in Node.js APIs. Open-source maintainers juggling reviews without full-time staff, or security-focused groups using it as a claude code security review GitHub action alongside human oversight. Ideal for repos already documenting conventions in CLAUDE.md.

Verdict

Solid for Claude fans wanting a claude PR review GitHub action—excellent docs, copy-paste workflows, and MIT license make it dead simple to try. With just 16 stars and 1.0% credibility score, it's early-stage and unproven at scale; test on non-critical repos first.

(187 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.