ClawHub
Back to skill

Security audit

Knowledge Ontology

Security checks across malware telemetry and agentic risk

Overview

This skill is a local knowledge-graph helper that writes and queries local ontology files, with no evidence of hidden exfiltration or destructive behavior.

Install this only if you want an agent to maintain a local structured memory store. Review commands before allowing create, migrate, compact, or plan operations, and do not put raw passwords or API keys into the ontology file.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (2)

Vague Triggers

Medium
Confidence
89% confidence
Finding
The skill includes broad operational language such as executing generic subsystem operations based on user input, without clearly limiting when the skill should activate. In an agent setting, vague activation criteria can cause over-invocation and unintended use of the skill's write/exec capabilities on prompts that only loosely match the topic.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill instructs the agent to run commands that create directories, touch files, append schema/data, and execute migrations, but it does not present a prominent warning that these actions modify local state. In an agent with exec access, this increases the risk of silent filesystem changes, accidental data corruption, or unsafe persistence triggered from natural-language requests.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.