red-agent

Q: How to install red-agent?

Create the agents directory: mkdir -p .claude/agents. Then add the config to .claude/agents/red-agent.md. Finally, @red-agent in Claude Code.

Q: What is red-agent best for?

red-agent is categorized under General. It covers: security, documentation, agent, plugin, red-team, adversarial, analysis, pr-review.

New

Community RegistryGeneralby Red Agent Team

Adversarial red/rainbow team analysis plugin for Claude Code. Finds weaknesses in LLM interactions through systematic probing using a 10x10 attack taxonomy (11 categories for PR analysis including code-duplication). Includes automated fix orchestrator with GitHub integration, PR analysis with 4 git modes, and comprehensive security documentation.

First seen 6/13/2026

Summary

The Red Agent skill systematically probes LLM interactions for security weaknesses using a 10x10 attack taxonomy, covering adversarial prompts, code-duplication, and PR analysis.

It includes an automated fix orchestrator with GitHub integration and comprehensive security documentation, making it essential for developers building robust, secure AI-powered applications.

Install & Usage

Create the agents directory

mkdir -p .claude/agents

Save the agent file

Add the configuration to .claude/agents/red-agent.md

Invoke with @agent-name

@red-agent

Use Cases

Identify vulnerabilities in your LLM-based application by running adversarial prompt attacks from a 10x10 taxonomy.

Automate PR reviews to detect security issues, code duplication, and potential weaknesses in code changes.

Generate comprehensive security documentation for your project based on red-team analysis results.

Integrate with GitHub to automatically fix identified vulnerabilities using the orchestrator.

Test your model's robustness against prompt injection, jailbreaking, and other adversarial techniques.

Analyze codebases for duplicate code and security anti-patterns using four different git modes.

Usage Examples

/red-agent analyze --taxonomy full --target my-app

/red-agent pr-review --repo owner/repo --pr 42 --mode deep

/red-agent fix --vuln-id VULN-001 --auto-commit

View source on GitHub

securitydocumentationagentpluginred-teamadversarialanalysispr-review

Security Audits

LicenseUnknownSourceWarnRepositoryPass

Frequently Asked Questions

What is red-agent?

The Red Agent skill systematically probes LLM interactions for security weaknesses using a 10x10 attack taxonomy, covering adversarial prompts, code-duplication, and PR analysis. It includes an automated fix orchestrator with GitHub integration and comprehensive security documentation, making it essential for developers building robust, secure AI-powered applications.

How to install red-agent?

To install red-agent: create the agents directory (mkdir -p .claude/agents), then add the config to .claude/agents/red-agent.md. Finally, @red-agent in Claude Code.

What is red-agent best for?

red-agent is a agent categorized under General. It is designed for: security, documentation, agent, plugin, red-team, adversarial, analysis, pr-review. Created by Red Agent Team.

What can I use red-agent for?

red-agent is useful for: Identify vulnerabilities in your LLM-based application by running adversarial prompt attacks from a 10x10 taxonomy.; Automate PR reviews to detect security issues, code duplication, and potential weaknesses in code changes.; Generate comprehensive security documentation for your project based on red-team analysis results.; Integrate with GitHub to automatically fix identified vulnerabilities using the orchestrator.; Test your model's robustness against prompt injection, jailbreaking, and other adversarial techniques.; Analyze codebases for duplicate code and security anti-patterns using four different git modes..