ClawHub
Back to skill

Security audit

Etherscan MCP

Security checks for vulnerabilities and agentic risk

Overview

The skill is a coherent Etherscan MCP helper with disclosed network and API-key use, and no evidence of hidden execution, exfiltration, or destructive behavior.

Install only if you are comfortable giving the agent access to an Etherscan API key and allowing it to create a UXC auth binding for the Etherscan MCP endpoint. Treat verifySourceCode as a submission action and confirm it explicitly before use.

A.I.G

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

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 (2)

Tp4

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding
There is a clear description-behavior mismatch. The declared purpose describes an operational skill for interacting with Etherscan MCP to investigate blockchain data. The supplied code instead validates that the skill's files and documentation contain required text and structure. While the script references Etherscan MCP commands and operations by name, it only searches for those strings in markdown/config files and never executes balance, transaction, token, or contract lookup actions. This is a materially different primary purpose, so it should be flagged as a mismatch.

Description-Behavior Mismatch

Medium
Confidence
93% confidence
Finding
The skill is framed as primarily read-only investigative tooling, but it exposes `verifySourceCode`, which is a write-like action against an external service. This can lead operators or downstream agents to invoke a state-changing or submission-oriented operation without sufficient confirmation, especially if they trust the skill's read-first positioning.

Static analysis

No suspicious patterns detected.