jinx

Security checks across malware telemetry and agentic risk

Overview

This is an instruction-only roundtable discussion mode with a broad trigger caveat but no evidence of data access, code execution, persistence, or destructive behavior.

Install this if you want a Chinese roundtable facilitation mode. Invoke it explicitly with phrases like “圆桌：<topic>” or “roundtable: <topic>”; be aware it may activate on loosely similar requests and will steer the assistant into a structured Chinese discussion format.

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 activation condition is overly broad because it applies not only to explicit trigger words but also to essentially any request that could benefit from 'multi-perspective deep discussion.' In a skill-routing system, this can cause unintended interception of general prompts, override more specific skills, and expand the prompt’s influence surface across unrelated topics, which is a prompt-scope security and safety issue even if not overtly malicious.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal