ClawHub

Advisory Council

Security checks across malware telemetry and agentic risk

Overview

This appears to be a market-analysis helper that runs a local script, with no artifact-backed evidence of theft, persistence, or destructive behavior.

Install only if you are comfortable with a skill that may execute a local Python market-analysis script. Review the script before use if available, and ask for confirmation before running analyses that could incur API costs or influence financial decisions.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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
89% confidence
Finding
The skill recommends triggering expensive, external-command execution for broad, natural-language requests such as general market questions or routine workflows. This can cause unintended runs, unnecessary API usage, and execution of a live script without clear user confirmation, which is risky because the skill explicitly instructs the agent to execute shell commands and trust the script's real output.

Natural-Language Policy Violations

Low
Confidence
82% confidence
Finding
The instruction to report results to 'Boss Man' imposes a specific address style without user opt-in, which can lead to unprofessional or inappropriate interactions and social-engineering style conditioning. In this skill, that concern is slightly elevated because the document also contains imperative language pushing real command execution, making the tone-control instruction part of a broader pattern of agent steering.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal