ClawHub

Crypto Trading

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed crypto-trading assistant, but it describes recurring automatic market-order trading without enough user control, safety limits, or data-sharing boundaries.

Review carefully before using with a live exchange account. Confirm the missing implementation, use trade-only keys with withdrawals disabled, set strict position and loss limits, understand what data is sent to external AI providers, and make sure there is a clear way to stop the 10-minute automation and remove stored trading state.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Missing User Warnings

High
Confidence
97% confidence
Finding
The skill describes an autonomous trading loop that executes market orders and updates persistent state, but it does not warn the user that it can place real trades without per-action confirmation. In the context of crypto trading, hidden autonomous execution is especially dangerous because market orders can incur immediate financial loss, slippage, and unintended repeated trades while file/state mutations make recovery and auditing harder.

Missing User Warnings

Medium
Confidence
94% confidence
Finding
The skill sends market analysis requests to external AI services while also referencing account holdings and balances, but it does not disclose what data may be transmitted off-system. This creates privacy and confidentiality risk because portfolio data, strategy signals, and potentially sensitive financial context could be shared with third-party providers without informed consent.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal