Question 1

What is Clawned?

Accepted Answer

Clawned is a security platform built specifically for the OpenClaw ecosystem. It offers a free public scanner that performs deep analysis of every OpenClaw skill — scanning the entire source code, scripts, and resources for 60+ threat patterns including reverse shells, credential theft, prompt injection, and supply chain attacks. It also provides a protection agent you install on your OpenClaw machine for real-time skill monitoring.

Question 2

What are OpenClaw skills and why do they need scanning?

Accepted Answer

OpenClaw skills are community-built packages that extend what your OpenClaw AI assistant can do — each containing source code, scripts, and configuration that run on your machine. ClawHub, the public registry, hosts 10,700+ skills. However, security research has found that over 36% of ClawHub skills contain security flaws, and campaigns like ClawHavoc have poisoned the registry with over 1,184 malicious skills that steal credentials, deploy stealers, and exfiltrate data.

Question 3

Do I need to sign up to scan OpenClaw skills?

Accepted Answer

No. The public scanner is completely free and requires no signup. Just paste a GitHub or ClawHub URL and get a full security report in under 10 seconds. You only need to sign in if you want the protection agent for continuous monitoring of your OpenClaw installation.

Question 4

How does Clawned protect against ClawHub supply chain attacks?

Accepted Answer

Clawned performs deep analysis of every skill's complete source code, bundled scripts, and resources for 60+ threat patterns. This includes detecting ClawHavoc-style attacks (fake prerequisites that install stealers), credential harvesting, prompt injection, reverse shells, and data exfiltration. Clawned has already blocked 230+ malicious ClawHub skills and protected 14,000+ developer machines.

Question 5

What is ClawHavoc and how does Clawned detect it?

Accepted Answer

ClawHavoc is a coordinated supply chain attack campaign that poisoned the ClawHub registry with over 1,184 malicious skills designed to steal credentials and deploy stealers. Clawned detects ClawHavoc-style attacks by analyzing the entire skill source code for fake prerequisites, obfuscated payloads, hidden shell commands, and credential harvesting patterns using 60+ specialized detection rules.

Question 6

Can I integrate Clawned into my CI/CD pipeline?

Accepted Answer

Yes. Clawned provides a REST API that you can integrate into any CI/CD pipeline. Generate an API key from your settings page, then use curl or any HTTP client to scan skills before deployment. This ensures no malicious OpenClaw skills enter your production environment.

Question 7

How do I install the Clawned protection agent?

Accepted Answer

Install Clawned with a single command: 'clawhub install clawned'. Then configure your API key with 'export CLAWNED_API_KEY=your-key' and run 'python3 ~/.openclaw/workspace/skills/clawned/scripts/agent.py daemon' to start continuous real-time monitoring of all your OpenClaw skills.

Question 8

Is Clawned compatible with the AgentSkills standard?

Accepted Answer

Yes. Clawned natively understands the AgentSkills open standard adopted by OpenClaw, Claude Code, GitHub Copilot, and Cursor. It parses SKILL.md frontmatter, audits install_spec dependencies, traces required_permissions, and analyzes every referenced script and resource file across the entire AgentSkills ecosystem.

Question 9

What types of threats does Clawned detect?

Accepted Answer

Clawned detects 60+ threat patterns across 10 categories: Remote Code Execution, Reverse Shells, Credential Harvesting, Data Exfiltration, Prompt Injection, Supply Chain Attacks (including ClawHavoc), Privilege Escalation, Obfuscation, Persistence Mechanisms, and Filesystem Abuse. It uses both static analysis and AI-powered detection to catch novel attack patterns.

OpenClaw & ClawHub Skill Security Reports