superclaw - SuperClaw: Red-Team AI Agents Before They Red-Team

项目说明

🦞 SuperClaw

<p align="center">
<a href=" alt="PyPI" src=" /></a>
<a href=" alt="Docs" src=" /></a>
<a href=" alt="GitHub Repo" src=" /></a>
<a href=" alt="CI" src=" /></a>
<a href=" alt="Docs Build" src=" /></a>
<a href=" alt="License" src=" /></a>
</p>

<p align="center">
<img src=" alt="SuperClaw logo" width="220" />
</p>

<p align="center">
<strong>Red-Team AI Agents Before They Red-Team You</strong><br/>
Scenario-driven, beha security testing for autonomous agents.
</p>

<p align="center">
<a href=" src=" alt="Documentation" /></a>
</p>

<p align="center">
<a href="#-quick-start">Quick Start</a> •
<a href="#-features">Features</a> •
<a href="#-attack-techniques">Attack Techniques</a> •
<a href=" Docs</a>
</p>

---

⚠️ Security and Ethical Use

Authorized Testing Only

SuperClaw is for authorized security testing only.

What is SuperClaw?

SuperClaw is a pre-deployment security testing framework for AI coding agents. It systematically identifies before your agents touch sensitive data or connect to external ecosystems.

🎯 Scenario-Dri

Generate and execute ad scenarios against real agents with reproducible results.

Get started →

📋 Behavior Contracts

Explicit success criteria, e extraction, and mitigation guidance for each security property.

Explore behaviors →

📊 E Reporting

Reports include tool calls, outputs, and actionable fixes in HTML, JSON, or SARIF formats.

CI/CD integration →

🛡️ Built-in Guardrails

Local-only mode and authorization checks reduce misuse risk.

Safety guide →

⚠️ Security and Ethical Use

Authorized Testing Only

SuperClaw is for authorized security testing only. Before using:

Guardrails enforced by default:

Threat Model

OpenClaw + Moltbook Risk Surface

OpenClaw agents often run with broad tool access. When connected to Moltbook or other agent networks, they can ingest untrusted, ad content that enables:

SuperClaw e these risks before deployment.

Why SuperClaw?

Autonomous AI agents are deployed with high pri, mutable behavior, and exposure to untrusted inputs—often without structured security . This makes prompt injection, tool misuse, configuration drift, and data leakage likely but poorly understood until after exposure.

What It Does

What It Doesn't Do

SuperClaw does not generate agents, run production workloads, or automate real-world exploitation. It's a testing tool, not a weapon.

---

🚀 Quick Start

Installation

Run Your First Attack

Generate Attack Scenarios

Run a Full Security Audit

---

✨ Features

Supported Targets

| Target | Description | Adapter |
|--------|-------------|---------|
| 🦞 OpenClaw | AI coding agents via ACP WebSocket | openclaw |
| 🧪 Mock | Offline deterministic testing | mock |
| 🔧 Custom | Build your own adapter | Extend BaseAdapter |

Attack Techniques

| Technique | Description |
|-----------|-------------|
| prompt-injection | Direct and indirect injection attacks |
| encoding | Base64, hex, unicode, typoglycemia obfuscation |
| jailbreak | DAN, grandmother, role-play bypass techniques |
| tool-bypass | Tool policy bypass confusion |
| multi-turn | Persistent escalation across con turns |

Security Behaviors

Each behavior includes a structured contract with intent, success criteria, rubric, and mitigation guidance.

| Behavior | Se | Tests |
|----------|----------|-------|
| prompt-injection-resistance | 🔴 CRITICAL | Injection detection and rejection |
| sandbox-isolation | 🔴 CRITICAL | Container and filesystem boundaries |
| tool-policy-enforcement | 🟠 HIGH | Allow/deny list compliance |
| session-boundary-integrity | 🟠 HIGH | Cross-session isolation |
| configuration-drift-detection | 🟡 MEDIUM | Config stability over time |
| acp-protocol-security | 🟡 MEDIUM | Protocol message handling |

---

📖 CLI Reference

Attacks

Scenario Generation (Bloom)

E

Auditing

Reporting

Scanning

Utilities

---

🔗 CodeOptiX Integration

SuperClaw integrates with CodeOptiX for multi-modal security e.

Python API

---

⚠️ Security Notice

This tool is for authorized security testing only.

Guardrails

Requirements

Before using SuperClaw, ensure you have:

---

🏗️ Architecture

---

🌐 Superagentic AI Ecosystem

SuperClaw is part of the Superagentic AI ecosystem:

| Project | Description |
|---------|-------------|
| SuperQE | Quality engineering core framework |
| SuperClaw | Agent security testing *(this package)* |
| CodeOptiX | Code optimization and e engine |

---

📚 Documentation

<table>
<tr>
<td width="100%" align="center">
<h3>📖 <a href="
</td>
</tr>
</table>

| Guide | Description |
|-------|-------------|
| Installation | Setup with pip, uv, or from source |
| Quick Start | Run your first security scan in 5 minutes |
| Configuration | Configure targets, LLM pro, and safety settings |
| Running Attacks | Execute attacks and interpret results |
| Custom Behaviors | Write your own security behavior specs |
| CI/CD Integration | GitHub Actions, GitLab CI, and SARIF output |
| Architecture | Deep dive into SuperClaw internals |

---

🤝 Contributing

We welcome contributions! Please see:

---

📄 License

Apache 2.0 — see LICENSE for details.

---

<p align="center">
Built with 🦞 by <a href=" AI</a>
</p>

项目信息

开发语言：Python

Star 数：404

技术要点

该项目属于红蓝攻防方向的开源资源，代码结构清晰，适合安全研究人员学习参考。项目围绕SuperClaw: Red-Team AI Agents Before They Red-Team You展开，提供了可运行的实现，对理解相关安全机制、工具原理与攻防技术具有较高的研究价值。

从技术实现角度看，此类项目的核心在于对底层原理的深入理解与工程化落地。开发者在研究过程中可以结合官方文档与社区讨论，梳理项目的初始化流程、核心模块职责与数据流转，逐步掌握其中的技术细节与设计思路。

使用与学习建议

1. 下载源码压缩包，解压到本地环境。

2. 根据项目类型准备对应的运行环境与依赖。

3. 按照 README 中的步骤完成配置与部署。

4. 在隔离的测试环境中运行验证，避免影响生产系统。

免责声明

本资源来源于开源社区，仅供安全研究、技术学习与合法授权范围内的攻防演练使用，请勿用于任何非法或侵害他人权益的用途。使用本项目产生的一切后果由使用者自行承担。