Security audit

Knowledge Mgmt Tool Free

Security checks across malware telemetry and agentic risk

Overview

This knowledge-management skill is not clearly malicious, but it asks for read, write, and command-running authority without enough limits or user-control guidance.

Install only if you are comfortable giving this skill permission to modify files and run local commands. Prefer using it in a restricted workspace, review each proposed file change or command before execution, and avoid granting it access to sensitive directories or credentials.

SkillSpector

By NVIDIA

Vulnerability Patterns

Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
Behavioral ASTexec() Call, eval() Call, Dynamic Import
MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands

Findings (5)

Description-Behavior Mismatch

Medium

Confidence: 95% confidence
Finding: The manifest presents a simple personal knowledge-management skill, yet it exposes exec capability and later classifies itself as MD+EXEC. That creates a privilege mismatch: ordinary content-management prompts could cause shell-command execution in an environment where users would not reasonably expect it.

Context-Inappropriate Capability

High

Confidence: 98% confidence
Finding: The declared command-execution capability is not justified by the stated purpose of auditing knowledge, generating templates, or organizing documents. Unnecessary execution rights materially expand the attack surface, enabling prompt-injection-driven system commands, file access, or chained abuse unrelated to the skill's business function.

Intent-Code Divergence

Medium

Confidence: 90% confidence
Finding: The documentation frames the skill as harmless Markdown-based usage with no extra API key, while also advertising executable-command capability. That mismatch can mislead users and operators into underestimating the risk, making unsafe deployment or automatic invocation more likely.

Vague Triggers

Medium

Confidence: 88% confidence
Finding: The example triggers are broad, ordinary natural-language requests such as auditing knowledge status or writing manuals. Broad triggers increase the chance the skill is invoked unintentionally during unrelated conversations, which is more dangerous here because the skill also has write/exec-capable tooling.

Missing User Warnings

Medium

Confidence: 97% confidence
Finding: The skill exposes write and exec capabilities without clearly warning users that it may alter the filesystem or run system commands. In an agent setting, missing warnings and consent boundaries can lead to unintended file modification, data loss, or execution of attacker-influenced commands.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.