Crypto Whale Alerts

• Outcome: Pass
• Latest audit: May 29, 2026 at 8:46 PM UTC
• Version:
  1.0.0

Overview

This is a local mock-data crypto alert helper with misleading trading-label bugs, but it does not show hidden access, credential use, persistence, or destructive behavior.

Install only as a local demo or starting point. Do not treat its output as live market intelligence or trading advice until the mock-data limitation and inflow/outflow classification are corrected and independently tested.

SkillSpector

By NVIDIA

Vulnerability Patterns

• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (2)

Intent-Code Divergence

Medium

• Confidence: 93% confidence
• Finding: The function claims to detect accumulation/distribution, but its implemented inflow/outflow semantics are inconsistent with the skill description. In a trading-alert skill, semantic inversion can cause downstream agents or users to interpret bullish activity as bearish, or vice versa, leading to incorrect automated decisions or financial losses.

Intent-Code Divergence

High

• Confidence: 98% confidence
• Finding: The alert labeling logic marks transactions as INFLOW when the sender is a known whale address, which contradicts the manifest's stated semantics that exchange receiving indicates inflow and outbound movement to cold storage indicates outflow. Because this skill produces actionable trading alerts, reversed labels can directly mislead operators or autonomous agents into taking the opposite market action.

VirusTotal

65/65 vendors flagged this skill as clean.