ClawHub

finchain-skill

Security checks across malware telemetry and agentic risk

Overview

This is an instruction-only FinChain/FUSD finance information skill with disclosed external lookups and no evidence of hidden access, unsafe actions, or credential use.

Reasonable to install for informational FinChain/FUSD assistance. Because it covers financial products and may use remote or live sources, verify yields, reserve evidence, risk ratings, and product availability on official sources before making financial decisions.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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
Findings (2)

Vague Triggers

Medium
Confidence
84% confidence
Finding
The manifest description is broad enough to activate the skill for generic finance-related prompts, including investment-style or stablecoin questions that may exceed the skill’s validated scope. Over-broad routing can cause the agent to answer with domain-specific authority in situations where other safer or more appropriate skills should handle the request, increasing the chance of misleading financial guidance or unintended use of external sources.

Natural-Language Policy Violations

Medium
Confidence
76% confidence
Finding
Forcing a default locale of zh-HK when the user’s language is ambiguous can cause the agent to respond in an unintended language, which may reduce user comprehension of financial or risk-related information. In a finance-oriented skill, misunderstanding product facts, yield caveats, or risk disclosures because of language mismatch can materially degrade safe use, even though this is not a direct code-execution issue.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal