SClawHub is a security scanner specifically designed for OpenClaw AI agent skills. Its primary purpose is to scan these skills for security vulnerabilities to prevent potential data theft or API key compromises, as OpenClaw agents have full system access, making security critical.

Key features include automated analysis combined with AI review to detect various security threats such as malware, credential theft, network calls, and code execution risks. The system provides a trust score on a scale of 0 to 100, with detailed reports that categorize skills into safe (scores 95-100), medium risk (scores 65-94), and critical risk (scores 0-64) levels.

Additionally, SClawHub is 100% open source, promoting transparency and allowing community audits. The scanner utilizes Semgrep for static code analysis and Claude AI for enhanced security review. A Chrome extension is available for easy browser-based access to scans, and the platform has scanned over 28 skills to date, with the number continuously updating.

The product works by allowing users to access skill URLs through the sclawhub.com domain instead of the original clawhub.ai. When a skill is accessed, it undergoes a comprehensive security scan, and the results are displayed with the trust score and report. The open-source approach ensures that the methodology is transparent and can be improved by contributors.

Benefits include enhanced security for OpenClaw users, enabling safe installation of skills and prevention of malicious code. Use cases involve verifying the safety of skills before integration into AI agents to avoid security breaches. The target audience includes developers and users of OpenClaw who need to ensure the security of their systems, especially when dealing with third-party skills.

Technical details include the use of Semgrep and Claude AI for scanning, with the project being MIT licensed and welcoming contributions from the community. The Chrome extension facilitates seamless integration for users browsing skills.