Knowledge Gaps

Security checks across malware telemetry and agentic risk

Overview

This skill has a reasonable purpose, but it tells the agent to automatically run an unbundled local Python script and persist unanswered user questions.

Review before installing. The skill should only be enabled if you intentionally want unanswered questions saved, and the referenced log-knowledge-gap.py script should be packaged, reviewed, and scoped before any agent follows the exec instruction.

SkillSpector

By NVIDIA

Vulnerability Patterns

Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
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

Findings (1)

Vague Triggers

Medium

Confidence: 89% confidence
Finding: The trigger condition is broad and subjective: any time Hans 'cannot answer' or says a stock phrase, the skill instructs it to execute a logging command. This can cause unintended execution frequency, logging of sensitive user prompts without adequate gating, and behavior that is easy for users to manipulate by eliciting uncertainty.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal