Security audit

Stablecoin Depeg Monitor

Security checks for vulnerabilities and agentic risk

Overview

This skill is a financial monitoring tool that advertises real-time stablecoin and arbitrage signals, but its commands generate simulated data without clearly labeling outputs as simulated.

Review carefully before installing. The script appears technically low-risk to your machine, but do not use its outputs for trading, treasury operations, compliance, or real depeg monitoring unless the data source is replaced with verified live market integrations and the tool clearly distinguishes simulation from production data.

SkillSpector

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 (4)

Description-Behavior Mismatch

Medium

Confidence: 98% confidence
Finding: The implementation presents itself as a real-time stablecoin monitor but only generates synthetic prices and liquidity data. In a trading, treasury, or DeFi risk-management context, this can cause operators to rely on fabricated signals, miss actual depegs, or act on false alerts, creating real financial and operational risk.

Description-Behavior Mismatch

Medium

Confidence: 97% confidence
Finding: The history command claims to provide historical stability analysis but instead fabricates a 30-day series. Users may make risk assessments, compliance decisions, or post-mortem analyses based on invented depeg events and stability scores, leading to materially incorrect conclusions.

Description-Behavior Mismatch

Medium

Confidence: 97% confidence
Finding: The arbitrage feature reports opportunities using simulated venue prices rather than real discrepancies. In this skill's trading-focused context, that is especially dangerous because users could execute trades or automate strategies on nonexistent spreads, causing direct financial losses.

Intent-Code Divergence

Low

Confidence: 94% confidence
Finding: The top-level documentation markets the tool as real-time on-chain and CEX monitoring even though the code is only a simulator. Misleading security- or finance-relevant documentation increases the chance of unsafe deployment and magnifies the harm of the simulated outputs.

Static analysis

No suspicious patterns detected.