ClawHub
Back to skill

Security audit

Gumroad

Security checks for vulnerabilities and agentic risk

Overview

The skill mostly matches its Gumroad-management purpose, but it under-labels refund and receipt-resend actions that can affect money or customers.

Install only if you are comfortable giving the connector access to your Gumroad account. Before any refund, shipment update, or receipt resend, require an explicit confirmation of the target sale and exact effect, even if the skill does not label the action as a write. Review the oo CLI installer source or use a safer installation path before running the one-line install commands.

SkillSpector

Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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
Findings (3)

Intent-Code Divergence

Medium
Confidence
97% confidence
Finding
The Safety section says only actions tagged [write] or [destructive] change state, but the listed untagged actions include operations such as refunding a sale and resending a receipt, both of which can alter user-visible state or trigger external effects. This can mislead an agent into executing side-effecting actions without confirmation, increasing the risk of unauthorized refunds or customer communications.

Vague Triggers

Medium
Confidence
89% confidence
Finding
The trigger description instructs the agent to use this skill for ANY Gumroad request instead of calling the API directly, which is overly broad and can cause the skill to be invoked in situations where a narrower or safer path would be more appropriate. In practice, broad auto-selection increases the chance of unintended state-changing operations being routed through this skill without sufficient task-specific safeguards.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The setup guidance includes shell and PowerShell one-liners that fetch and execute remote scripts directly, but it does not provide an explicit warning about the trust and supply-chain risk of piping network content into a shell. If followed automatically or casually, this could lead to arbitrary code execution on the user's machine if the distribution endpoint or transport is compromised.

Static analysis

No suspicious patterns detected.