Anthropic says its Claude Security plugin for Claude Code is now in beta and available to all Claude Code users, adding terminal-based workflows to scan repositories, review code changes, validate findings, and draft security patches.
Anthropic has expanded access to its Claude Security plugin for Claude Code, making the beta available to all Claude Code users.
Anthropic’s Claude Security product page says the plugin is designed to bring security review workflows directly into Claude Code. According to Anthropic, users can scan code, validate suspected issues, and create patches from within the Claude Code environment.
The company’s Claude Code documentation describes the plugin as a way to check either an entire repository or a narrower set of changes. That includes scans against a branch diff, pull request diff, or a specific commit. The documentation also says the plugin can generate patches for user review, positioning it as a developer-facing security tool rather than a standalone scanner outside the coding workflow.
Anthropic’s Claude Code docs say the Claude Security plugin supports full-repository scans as well as targeted scans of recent changes. The targeted options matter for teams that want to review security risks before code is merged, since scanning only a diff can reduce noise compared with checking an entire codebase every time.
The docs describe use cases such as scanning changes before committing and running a broader scan across a codebase from the terminal. Anthropic’s product page frames the workflow around three steps: scan, validate, and patch. The company says these steps happen inside Claude Code, the command-line coding tool that connects developers to Claude while they work.
Anthropic does not present the plugin as a replacement for human review or established security processes in the cited documentation. The docs say generated patches are for review, which implies developers remain responsible for checking proposed fixes before applying or merging them.
Anthropic’s Claude Code plugin documentation says users can install prebuilt plugins through Claude Code marketplaces. The official marketplace is identified as claude-plugins-official.
For Claude Security specifically, Anthropic’s Claude Code docs list the installation command as:
/plugin install claude-security@claude-plugins-official
The broader marketplace documentation confirms the same installation pattern for plugins from the official marketplace. This means the Claude Security plugin is distributed through Anthropic’s plugin system rather than as a separate download described in the cited sources.
The notable change is availability. Anthropic’s Claude Security product page states that the Claude Security plugin for Claude Code is now in beta and available for all Claude Code users. That widens access to a feature set focused on finding and fixing potential vulnerabilities while developers are already working in the terminal.
The company’s documentation emphasizes repository-wide scans, diff-based scans, validation, and patch generation. Those capabilities place the plugin in the growing category of AI-assisted developer security tools, but the cited Anthropic materials do not provide independent benchmark results or comparative accuracy claims.
For now, the most concrete information comes from Anthropic’s own product and documentation pages: Claude Code users can install the beta plugin from the official marketplace, run security scans against a repository or selected changes, and review generated patches inside Claude Code.
Anthropic has expanded access to its Claude Security plugin for Claude Code, making the beta available to all Claude Code users.
Claude Security arrives as a Claude Code plugin Anthropic’s Claude Security product page says the plugin is designed to bring security review workflows directly into Claude Code.
According to Anthropic, users can scan code, validate suspected issues, and create patches from within the Claude Code environment.
Continue reading