Anthropic's official X account, @claudeai, shared the following announcement.
"The Claude Security plugin for Claude Code is now available in beta. Scan your changes for vulnerabilities before you commit, or run a full scan across your codebase, all from your terminal on the Claude inference you already run."
In short, the Claude Security plugin for Claude Code has entered beta (beta — a trial stage before full release), letting developers scan their changes for vulnerabilities (vulnerability — a security flaw that could be exploited) before committing (commit — recording code changes to a repository), or run a full scan across an entire codebase. According to the post, both capabilities run entirely from the terminal, using the same Claude inference (inference — the computation an AI model performs to generate a response) the user already relies on.
What the Announcement Means
The core idea here is bringing security checks directly into the Claude Code workflow instead of requiring a separate external tool. Developers can check just their changed code before committing, or scan an entire repository when needed. The post emphasizes that both modes run on the user's existing Claude inference, without extra infrastructure or separate model calls.
What This Means for Users
This plugin creates an opportunity to catch security issues earlier, before a formal code review. However, details such as supported languages, the exact scope of scanning, pricing, usage limits, and a firm general availability date were not included in the official announcement. As a beta feature, its capabilities and terms may change going forward.
For the latest status and usage details, check Anthropic's official channels.