Anthropic Launches Claude Security Plugin for Claude Code in Beta: Multi-Agent Vulnerability Scanner in the Terminal

What's new

• Claude Security Plugin (beta): multi-agent vulnerability scanner integrated directly into Claude Code
• Inline detection: identifies potential security issues during development — before code reaches a PR or commit
• Multi-agent architecture: spawns multiple sub-agents for parallel examination of different parts of the codebase
• Terminal-native: works directly in the terminal without a separate CI/CD tool or web UI
• Suggested patches: in addition to detection, proposes concrete fixes for identified vulnerabilities

Why it matters

This is the first native security tool integrated directly into an AI coding assistant (not as an external scanner). Developers receive security feedback in real time while writing code — not after a code review or SAST scan in CI.

How to try it

The plugin is available in beta for Claude Code users. Activate via the /security command or through the plugin settings in Claude Code.