ClawHub
Back to skill

Security audit

Learning Hardened

Security checks across malware telemetry and agentic risk

Overview

This is a text-only learning preference skill that may persist teaching-style preferences, but it does not include code, installers, credentials, or unrelated capabilities.

Install this only if you are comfortable with the agent maintaining a persistent learning-preference profile. Before using it, confirm that new entries require your approval and that you can view, edit, and delete stored preferences; avoid allowing it to record medical, demographic, or identity details.

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)

Vague Triggers

Medium
Confidence
87% confidence
Finding
The skill description promises broad, always-on adaptation ('Auto-learns how you learn best') without clear trigger conditions, user consent flow, or limits on when profiling occurs. In an agent setting, this can cause the skill to activate across many contexts and accumulate persistent user-profile data more often than users would reasonably expect.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The file explicitly instructs the agent to 'observe and fill' preference sections and to detect patterns from successful explanations, which implies persistent profiling of user behavior. Although the guardrails exclude some sensitive categories, the skill still lacks a clear warning, consent mechanism, and transparency around retention, making silent profile writes the primary risk.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.