Dupont Analysis
Security checks across malware telemetry and agentic risk
Overview
The artifact is a coherent ClawHub/OpenClaw registry codebase with disclosed maintainer, CLI, moderation, and development workflows, and I did not find evidence of hidden or deceptive behavior.
Install or use this only in a ClawHub development or maintainer context. Review the maintainer skills and CLI commands before running them, keep API and deployment tokens scoped, and avoid using moderation or proof-publishing commands unless you intend to make authenticated changes to ClawHub or GitHub.
SkillSpector
By NVIDIA
Vulnerability Patterns
- Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
- Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
- Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
- Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
- Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
VirusTotal
66/66 vendors flagged this skill as clean.