Security audit
Cross-Disciplinary Creativity
Security checks across malware telemetry and agentic risk
Overview
The reviewed skills are high-trust maintainer and Convex workflow helpers, but their sensitive actions are disclosed, scoped, and gated by user confirmation.
Install only in an environment where you intend agents to help with ClawHub maintainer or Convex development work. Several workflows can make real production, moderation, email, or repository changes, so review generated commands carefully and keep the required confirmation/dry-run gates in place.
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
64/64 vendors flagged this skill as clean.
Static analysis
No suspicious patterns detected.