Claude-security — Claude plugin: security scans and analysis
An Anthropic plugin that adds security workflows to Claude: scanning, analysis and incident investigation inside the chat.
Open on GitHubWhat codex-security does
Claude-security is a plugin for Claude and Claude that bundles security workflows: scans, code analysis and digging into suspicious spots. It comes from Anthropic, repo openai/plugins.
Inside are ready-made flows for diff review, finding vulnerabilities and threat modeling. It installs like any other Claude plugin, no separate scripts to run.
The license is proprietary, so forking and editing the source is off the table. If you need an open tool you can bend to your own process, that is a reason to look elsewhere.
What it helps with
- run a security pass on a diff before merge
- find vulnerable spots in fresh code
- investigate an incident and figure out what happened
- build a threat model for a service
- check someone else's pull request for risky changes
Example prompts for the agent
- check this diff for security issues
- scan the repo and list the vulnerabilities
- walk through this incident and tell me what to fix
Questions about codex-security
How do I install Claude-security in Claude?
It installs from the openai/plugins repository through Claude's standard plugin mechanism. Check the Claude docs for the exact install commands.
What does Claude-security do?
It adds security workflows to Claude: code scanning, analysis and incident investigation. It works through ready-made flows rather than separate scripts.
Can I modify Claude-security for my own needs?
The license is proprietary, so editing the source or shipping your own version is not allowed. For custom work, pick an open tool instead.
