Anthropic is bringing the deep scanning capabilities of Claude Security to a wider audience with a new Claude Code plugin. The plugin is now in beta and available to all paid Claude Code users. Meanwhile, the managed version of Claude Security remains strictly enterprise-only.
With the plugin, users can scan entire codebases or narrow their focus to branch diffs, PR diffs, or individual commits. When a vulnerability is found, Claude can generate a patch, which users can review and apply at their discretion.
Claude Security isn't a brand new product — it entered public beta for enterprise customers on April 30, 2026, after a limited research preview under the name Claude Code Security.
The new plugin runs inside a Claude Code session, where multiple Claude agents work together to map out architecture, build threat models, and hunt for vulnerabilities. Every finding and patch gets reviewed by a separate agent before it's surfaced. Scanning activity counts against a user's Claude Code usage.