[Claude Security Decentralizes Claude Code: Individual Paid Users Can Also Scan Entire Repositories for Vulnerabilities] According to monitoring by Beating, Anthropic has developed the deep scanning capabilities of Claude Security into a Claude Code plugin. The plugin is now available in beta, and all Claude Code paid users can install it. The hosted version of Claude Security remains exclusive to Enterprise customers. Users can scan the entire code repository or focus on branch differences, PR differences, or single commits. After identifying vulnerabilities, users can choose to have it generate a patch, which they can then review and apply themselves. Claude Security itself is not a new product. It entered public testing for Enterprise customers on April 30, 2026, after previously being available in a limited research preview under the name Claude Code Security. The new plugin operates within Claude Code sessions. Multiple Claude Agents will analyze the architecture, build threat models, and identify vulnerabilities. Each finding and patch will be independently reviewed by another Agent. Scanning usage will count toward Claude Code consumption. [Original Link]