striver

Security checks across malware telemetry and agentic risk

Overview

This is an instruction-only persistence skill, but it broadly pushes the agent to keep acting and use tools without clear limits or normal stop points.

Install only if you explicitly want a strong persistence mode. Set clear boundaries that normal safety policies, user consent, clarification, refusal, and stopping still take precedence, especially before searching the web, reading private files, running commands, or making changes.

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

High

Confidence: 96% confidence
Finding: The trigger conditions are extremely broad and behavior-based, such as when progress 'stalls,' the agent is 'about to give up,' or 'passive behavior appears.' This can cause the skill to activate during ordinary interactions and override normal agent safeguards, pushing the agent into persistent action-seeking behavior even when clarification, refusal, or stopping would be safer.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal