Observer Effect Probe
Security checks across malware telemetry and agentic risk
Overview
This skill is a defensive, documentation-only probe description with no artifact-backed malicious behavior or hidden execution path.
Review any commands the agent proposes when actually probing another skill, especially uses of curl or python3, and run probes only in isolated environments without real credentials. Based on the supplied and inspected artifacts, there is no concrete reason to hold this skill for review.
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
65/65 vendors flagged this skill as clean.