2 min read
Add as a preferred source on Google

Anthropic Extends Claude Security to Track Code Changes and Fixes

Projects can connect up to eight repositories. Eligible Claude Enterprise organizations can receive monthly usage credits.

Anonymous hands update code on a laptop in a quiet lab / TokenPost.ai
Anonymous hands update code on a laptop in a quiet lab / TokenPost.ai

Anthropic has expanded Claude Security’s public beta with recurring scans of changed code and automatic fix tracking, giving enterprise security teams a more current view of vulnerabilities across connected repositories.

The Oct. 21 update moves the service beyond scans that users start or schedule. A new project begins with a deep scan. After the initial scan, the tool checks for new commits and reviews changed code regularly, with work shaped by risk and complexity.

A project can connect up to eight repositories, allowing teams to assess findings across related services. Teams can add deployment details and security policies to help the tool rate findings and explain how an attacker might reach vulnerable code.

When developers change code, Claude Security revisits open findings. It marks an issue fixed if it detects a remedy and can reopen the finding if the problem returns. Teams can bring findings into their workflows through an API, webhooks and a connector for Claude and Claude Code.

Claude Security entered public beta for Claude Enterprise customers on April 30. Its initial version offered scheduled and targeted scans, vulnerability findings and proposed patches for human review. The enterprise service is separate from Anthropic’s scanner for open-source projects.

“Now it keeps up as our code changes, and our team fixes issues right in Claude instead of switching to another tool,” said Jake Meredith of Gusto’s product security team.

Existing projects continue in the earlier version unless they are migrated. Migration creates a new project, copies its findings and archives the original. Members, webhooks, excluded directories, scan history and scheduled scans do not transfer, so teams seeking to retain those settings must configure them in the new project.

The updated service remains in public beta for Claude Enterprise customers. Usage is billed at direct token cost, with no additional platform fee. Eligible Claude Enterprise organizations can receive up to $25,000 in credits per month for two consecutive months. The offer ends at 7 p.m. ET on Dec. 31 (midnight Jan. 1, 2027 UTC), even if an organization has not used both full months.

Simon Yoon

Reporter

Simon Yoon reports on blockchain technology for TokenPost. Send corrections or tips to info@tokenpost.com.

Loading…