Knowledge Base Manager

Security checks across malware telemetry and agentic risk

Overview

This instruction-only skill coherently manages a local Markdown knowledge base, with disclosed local file creation scoped to a `knowledge/` folder.

Use it in a dedicated workspace or `knowledge/` directory, review generated paths before large reorganizations, and avoid saving secrets or reusing stored prompt/instruction entries without checking their source.

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 (2)

Missing User Warnings

Low
Confidence
90% confidence
Finding
The skill explicitly instructs initialization steps that create directories and copy files into the local filesystem, but it does not require an explicit confirmation or user-visible warning before making those changes. In an agent setting, filesystem-writing behavior can cause unintended modifications, especially if the user expected planning or dry-run behavior rather than execution.

Vague Triggers

Medium
Confidence
86% confidence
Finding
The text is a generic imperative request to save an article and perform classification, which can match or activate the skill without clear scoping, trust boundaries, or user confirmation. In a knowledge-base management skill, broad intake phrasing increases the chance that unintended or attacker-supplied content is processed, classified, and persisted, potentially enabling prompt-injection-style misuse or unwanted file creation.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal