ClawHub
Back to skill

Security audit

Equity Analyst Test

Security checks across malware telemetry and agentic risk

Overview

This stock-analysis skill appears coherent and non-malicious, but its investment-style scores and stderr logging deserve user caution.

Install only if you are comfortable with a tool that analyzes public Korean stock data and produces investment-style scores. Do not treat its output as financial advice or as a sole basis for trading decisions, and be aware that the current script may write full analysis details to stderr where orchestration logs could capture them.

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)

Natural-Language Policy Violations

Medium
Confidence
92% confidence
Finding
The skill metadata and examples strongly assume Korean-language interaction and do not offer a language choice or fallback, which can exclude or confuse users who requested analysis in another language. In a finance-related skill, language mismatch increases the chance that users misunderstand scoring, caveats, or recommendation labels, which can materially affect decision-making.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill generates investment-attractiveness scores and BUY/HOLD/AVOID-style verdicts but does not present a clear upfront warning that the output may influence real financial decisions. Although it says 'Do NOT give investment advice' in the output section, the structured recommendation framework still functions as actionable investment guidance, which can mislead users into treating it as professional advice without appropriate caution.

Missing User Warnings

Medium
Confidence
94% confidence
Finding
The script always emits the full structured analysis result to stderr after execution, even when the user only requested plain-text output. In environments where stderr is centrally logged, captured by orchestration systems, or exposed to other tenants, this can leak potentially sensitive input-derived financial analysis data without consent or clear user awareness.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.