Enterprise-Grade Security Scanning Lands in Claude Code
Professional code security scanning, once the exclusive domain of large engineering teams with enterprise budgets, is now accessible to individual developers. Anthropic has released a beta plugin for Claude Code that packages the core deep scanning intelligence from its Claude Security product, making it available to all paying users of the code assistant.
Unpacking the Scanning Capabilities
The new plugin offers granular control over what gets analyzed. Developers can initiate scans tailored to their immediate workflow needs directly within their coding environment.
- Full Repository Analysis: Conduct a comprehensive audit of an entire connected codebase to map architecture and identify systemic risks.
- Targeted Diff Scanning: Focus the analysis precisely on changes in a feature branch, an open Pull Request, or even a single commit to catch issues early.
- Integrated Remediation： When a vulnerability is detected, the plugin can generate a suggested fix or patch. Crucially, each finding and its corresponding patch are reviewed by a separate agent for validation, with the final decision to apply the changes resting entirely with the developer.
How It Works and What It Costs
The security scanning operates inside the Claude Code session using a multi-agent approach. Different specialized agents work in concert to understand the codebase structure, build a threat model, and then hunt for specific vulnerabilities, aiming for higher accuracy.
In terms of pricing, the computational resources used for security scans are billed as part of the user's existing Claude Code usage quota. There is no separate fee for activating the security features, lowering the barrier to adoption.
It's important to distinguish this new plugin functionality from the full, standalone Claude Security platform. The managed Claude Security service, which includes centralized dashboards and administrative controls, remains exclusively available to enterprise customers.