Security audit
persistent-skill-memory
Security checks across malware telemetry and agentic risk
Overview
The skill is transparent about its purpose, but it asks the agent to persistently modify its durable prompt and hook future skill changes without clear rollback or user-control boundaries.
Review this carefully before installing. It appears intended to solve skill recall, not to steal data or run hidden code, but it would guide your agent to persistently alter its durable prompt and hook future skill-related workflows. Only use it if you are comfortable with that persistence and have a clear way to inspect, disable, and remove the injected prompt block and hooks.
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.