Security audit
Loss Aversion and Prospect Theory
Security checks across malware telemetry and agentic risk
Overview
The artifacts describe disclosed ClawHub and Convex operational skills with powerful but user-gated admin actions and no hidden or deceptive behavior found.
Install only in a ClawHub/Convex operator environment where you are comfortable with staff-level CLI actions. Review commands before approving writes, especially user bans, role changes, package transfers, production migrations, and outbound email.
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.