Urgent.News

What's breaking now, across thousands of outlets.

AI

Claude Code Review Is GA and Never Fails a PR. Anthropic's Own Repos Block on Findings

Book: Observability for LLM Applications Also by me: Thinking in Go (2-book series) — Complete Guide to Go Programming + Hexagonal Architecture in Go My project: Hermes IDE | GitHub — an IDE for developers who ship with Claude Code and other AI coding tools Me: xgabriel.com | GitHub Anthropic won't merge a pull request to the Claude Code repos while a review finding is still open. Code Review for…

On October 15, Anthropic made its Claude Code review tool generally available to Claude Team and Enterprise organizations. The review system is designed to never fail a pull request check. When a review finding remains open, Anthropic's own repositories block the merge of the pull request. The review process involves a few agents reading the code diff and surrounding context, with a verification pass discarding any false positives.

The tool marks findings as Important, Nit, or Pre-existing, and posts them as inline comments on the pull request along with a dedicated check run named Claude Code Review. The review runs on Anthropic's infrastructure with read-only access to the code and only works with GitHub repositories. It is not available for organizations with Zero Data Retention, HIPAA configurations, or customer-managed encryption keys.

The team posts that a review never approves or blocks a pull request, and the check run always concludes as not a failure. The team also explains that approving a pull request is still a human decision, and the product posts findings, not approvals. If you want findings to block a merge, two options are available: requiring resolved conversations or gating on Important findings only. A pre-merge step script is provided to enforce these rules in your repository.

Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Read the original at dev.to →

More in AI

I Had 5 AI Features That Worked Individually. The Product Still Felt Broken

I Had 5 AI Features That Worked Individually. The Product Still Felt Broken. I had five AI features that all worked on their own.

  • Five AI features functioned independently, each working correctly.
  • Product felt incomplete despite individual feature successes.
  • Transition between features caused user confusion and frustration.

ZeroTrace: An Air-Gapped, Zero-Day Behavioral SOC Copilot

This is a submission for the Hacktoberfest Open-Source AI Challenge Week 2: Side Hustle What I Built Modern cybersecurity threats are evolving rapidly, but advanced security monitoring and Endpoint…

  • ZeroTrace is an AI-powered, local-first cybersecurity tool
  • Focuses on behavioral threat analysis and AI-assisted triage
  • Offers per-server subscriptions and managed incident triage services

More from Sunday 18 October →