ClawHub
Back to skill

Security audit

OnChat

Security checks across malware telemetry and agentic risk

Overview

The skill appears purpose-aligned, but it can let an agent use a wallet private key to spend ETH and post permanent public blockchain messages without clear per-action limits.

Install only with a dedicated low-balance Base wallet, never a primary wallet key. Set explicit rules for approved channels, maximum messages, maximum ETH spend, monitoring duration, and whether every outgoing message must be reviewed before submission. Do not allow the agent to post secrets, personal data, credentials, or wallet details because messages are public and permanent.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)

Lp3

Medium
Category
MCP Least Privilege
Confidence
91% confidence
Finding
The skill instructs operators to provide `ONCHAT_PRIVATE_KEY`, but the manifest does not declare environment-variable or secret access permissions. This creates a transparency and consent gap: an agent platform or user may not realize the skill expects sensitive credentials, increasing the risk of unintended secret exposure or use in write-capable blockchain actions.

Missing User Warnings

Medium
Confidence
94% confidence
Finding
The send command performs an implicit state-changing blockchain action by auto-joining the channel before sending the message if the wallet is not already a member. In an agent skill context, this is dangerous because a caller expecting one transaction may unknowingly authorize two on-chain actions, increasing gas spend and causing unintended membership/state changes without explicit consent.

Unpinned Dependencies

Low
Category
Supply Chain
Content
"private": true,
  "type": "module",
  "dependencies": {
    "tsx": "^4.19.2",
    "viem": "^2.23.2"
  }
}
Confidence
90% confidence
Finding
"tsx": "^4.19.2"

Unpinned Dependencies

Low
Category
Supply Chain
Content
"type": "module",
  "dependencies": {
    "tsx": "^4.19.2",
    "viem": "^2.23.2"
  }
}
Confidence
90% confidence
Finding
"viem": "^2.23.2"

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.