ClawHub
Back to skill

Security audit

Accounting And Finance Free

Security checks across malware telemetry and agentic risk

Overview

This is a read-only finance analysis skill with one minor troubleshooting instruction that should be treated cautiously but does not justify Review by itself.

Installers should understand this as a read-only financial analysis prompt. Do not allow automatic shell or network diagnostics from this skill; if network troubleshooting is needed, review and approve any command separately.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • 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 (3)

Intent-Code Divergence

Medium
Confidence
94% confidence
Finding
The skill explicitly declares itself as pure Markdown with no command-execution capability, yet later documentation instructs the user/agent to run `ping` for troubleshooting. This mismatch weakens trust boundaries: an agent or user may assume the skill never causes system interaction, but the instructions introduce real command execution behavior that was not transparently disclosed.

Intent-Code Divergence

Medium
Confidence
97% confidence
Finding
The error-handling section directs execution of `ping` commands despite the skill being labeled as Markdown-only. Even though `ping` is a common diagnostic tool, instructing command execution from within a supposedly non-executing skill can trigger unintended host/network interaction, create policy bypass opportunities, and normalize unsafe escalation from documentation to shell actions.

Missing User Warnings

Low
Confidence
92% confidence
Finding
The skill instructs users to run a network diagnostic command without clear warning that this causes system-level command execution and outbound network activity. In context, this is more concerning because the rest of the skill is framed as safe, read-only financial analysis, so users may not expect any operational actions at all.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.