Ensure Docs
Security checks across malware telemetry and agentic risk
Overview
The skills are mostly coherent, but the autoreview helper defaults to a broad sandbox/approval bypass that users should review before installing or running.
Install only if you trust the publisher and intend to use these workflows in a ClawHub/Convex development context. Before running autoreview, prefer `--no-yolo` or set the opt-out environment variable unless full local access is truly needed, and be aware that fallback reviewers may receive diff contents. Use moderation commands only with the exact target, reason, and confirmation you intend.
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
57/57 vendors flagged this skill as clean.