ClawHub
Back to skill

Security audit

ERCData

Security checks across malware telemetry and agentic risk

Overview

This skill is a coherent blockchain CLI for storing and verifying ERCData records, but users should treat its commands as real Base mainnet transactions and protect wallet keys carefully.

Install only if you intend to let an agent prepare or send Base mainnet ERCData transactions. Use a dedicated low-balance wallet, prefer protected environment/secret handling over passing --key on the command line, verify the contract and RPC endpoint, manually review write/access/snapshot actions, and store hashes or client-side encrypted payloads rather than raw sensitive data.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Tp4

High
Category
MCP Tool Poisoning
Confidence
84% confidence
Finding
The skill description overstates or misstates supported behavior, including claims around batch operations and encrypted storage that are not actually implemented, while underlying interfaces may expose additional privileged role-management functions. For a blockchain skill that handles private keys, access control, and on-chain data publication, this mismatch can cause operators or agents to make unsafe assumptions about privacy, available safeguards, and what transactions may be possible.

Missing User Warnings

Medium
Confidence
88% confidence
Finding
Accepting a private key via a command-line flag exposes the secret to shell history, process listings, audit logs, and orchestration telemetry. In an agent skill context, this is more dangerous because automated systems may pass arguments through observable wrappers, increasing the chance of credential disclosure and wallet compromise.

Missing User Warnings

Medium
Confidence
86% confidence
Finding
The store path sends user-supplied data directly to a blockchain transaction without any explicit warning that the content may become permanently replicated across network infrastructure. In this skill's context, users may assume 'private' means confidential, so lack of disclosure materially increases the risk of irreversible leakage of sensitive AI data, prompts, metadata, or personal information.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.