Self Verify

Security checks across malware telemetry and agentic risk

Overview

This is a small self-check prompt skill that may be a bit broad, but its behavior is disclosed and aligned with improving answer accuracy.

Install this if you want the agent to self-check important claims before answering. Expect it may make responses more cautious or slower, and review any configured search or memory tools because the skill encourages using them for verification.

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: 91% confidence
Finding: The skill activates on broad triggers like 'about to return a consequential answer' and 'any high-confidence claim,' which are subjective and likely to fire in many normal interactions. Overly broad invocation can cause unintended takeover of workflows, prompt interference, or denial-of-service style behavior where the agent repeatedly self-checks instead of completing the user task.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal