Security audit
Databox
Security checks across malware telemetry and agentic risk
Overview
The skill is mostly coherent for Databox automation, but one data-changing Databox action is not marked for confirmation even though the safety rules say untagged actions are safe to run directly.
Review this before installing if your Databox account contains business-critical metrics. Treat `push_dataset_data` as a write action and require explicit confirmation of the target dataset and payload, just like the skill already requires for create and delete actions.
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
63/63 vendors flagged this skill as clean.
Static analysis
No suspicious patterns detected.