agent-red-team-prompts

Q: How to install agent-red-team-prompts?

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

Q: What is agent-red-team-prompts best for?

agent-red-team-prompts is categorized under General. It covers: agent.

New

GitHub TrendingGeneralby ban10yuu

Copyable red-team scenarios for AGENTS.md, CLAUDE.md, hooks, skills, and agentic repos.

First seen 6/16/2026

Summary

This skill provides a library of copyable red-team scenarios designed to test and improve the security and reliability of agentic AI systems.

md, hooks, skills, and agentic repos by simulating adversarial prompts and behaviors.

Install & Usage

Create the skills directory

mkdir -p .claude/skills

Download the skill file

Add the configuration to .claude/skills/agent-red-team-prompts.md

Invoke in Claude Code

/agent-red-team-prompts

Use Cases

Test your AGENTS.md configuration against prompt injection attacks that attempt to override system instructions.

Validate that CLAUDE.md constraints are enforced when an agent is asked to perform disallowed actions.

Identify weaknesses in custom hooks by simulating malicious input designed to bypass validation logic.

Audit a skill's behavior under adversarial conditions, such as attempts to leak sensitive data or execute unauthorized commands.

Stress-test agentic repos with multi-step red-team scenarios that chain together several attack vectors.

Evaluate the robustness of your agent's tool-use permissions when faced with deceptive requests.

Usage Examples

/agent-red-team-prompts list all scenarios for AGENTS.md

Run the 'prompt injection' red-team scenario against my current project's AGENTS.md.

Show me a red-team scenario that tests if my CLAUDE.md can be overridden by a user message.

View source on GitHub

agent

Security Audits

LicenseUnknownSourceWarnRepositoryPass

Frequently Asked Questions

What is agent-red-team-prompts?

This skill provides a library of copyable red-team scenarios designed to test and improve the security and reliability of agentic AI systems. It helps developers identify vulnerabilities in AGENTS.md, CLAUDE.md, hooks, skills, and agentic repos by simulating adversarial prompts and behaviors.

How to install agent-red-team-prompts?

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

What is agent-red-team-prompts best for?

agent-red-team-prompts is a community categorized under General. It is designed for: agent. Created by ban10yuu.

What can I use agent-red-team-prompts for?

agent-red-team-prompts is useful for: Test your AGENTS.md configuration against prompt injection attacks that attempt to override system instructions.; Validate that CLAUDE.md constraints are enforced when an agent is asked to perform disallowed actions.; Identify weaknesses in custom hooks by simulating malicious input designed to bypass validation logic.; Audit a skill's behavior under adversarial conditions, such as attempts to leak sensitive data or execute unauthorized commands.; Stress-test agentic repos with multi-step red-team scenarios that chain together several attack vectors.; Evaluate the robustness of your agent's tool-use permissions when faced with deceptive requests..