ClawHub
Back to skill

Security audit

GMGN Skill Market

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed GMGN market-data helper, but users should be aware it can install an external CLI, configure an API key, and run limited network troubleshooting.

Install this only if you are comfortable using the external gmgn-cli package and storing a GMGN API key locally. During first-time setup, confirm before generating keys or saving credentials, and remove /tmp/gmgn_private.pem if it is no longer needed. If troubleshooting runs, expect it to inspect IPv6 interface data and contact an IP-check service. Treat token signals as research, not financial advice.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (3)

Context-Inappropriate Capability

Medium
Confidence
96% confidence
Finding
The skill instructs the agent to run host-level network diagnostic commands (`ifconfig`, `ip addr`, and an external IP check) that are not necessary for market-data retrieval and expose system/network metadata. In an agent environment, this broadens the skill from API usage into local environment reconnaissance, which can leak sensitive infrastructure details or normalize unrelated command execution when API auth fails.

Context-Inappropriate Capability

Medium
Confidence
94% confidence
Finding
The setup flow directs the agent to generate cryptographic key material on the local system and manage onboarding for an external service, which exceeds the skill's stated purpose of retrieving market data. Even though it generates an Ed25519 keypair for API setup, having a general-purpose skill create keys and handle auth bootstrap increases the chance of secret mishandling, persistence of sensitive material, or unintended reuse of local cryptographic assets.

Vague Triggers

Medium
Confidence
85% confidence
Finding
The trigger phrases are broad enough to match common conversational requests like 'what's pumping' or 'hot coins,' which can cause the skill to activate in more contexts than intended. Over-broad routing is risky because this skill contains side-effecting setup and diagnostic instructions, so accidental invocation could lead the agent toward unnecessary command execution or credential handling.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.