ClawHub
Back to skill

Security audit

ClawPlot

Security checks for vulnerabilities and agentic risk

Overview

The skill is for a real paid ordering workflow and is mostly coherent, but it can send shipping details and order data to an external service without clear consent or narrowly scoped invocation guidance.

Review before installing. This skill should only be used when you explicitly want to order physical artwork from ClawPlot and are comfortable sending the SVG, shipping name/address, and payment-method choice to clawplot.com. Confirm price, recipient, address, and payment flow before allowing an agent to submit an order.

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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (4)

Vague Triggers

Medium
Confidence
95% confidence
Finding
The skill description uses broad triggers like 'make something real,' 'commission artwork,' and 'give a meaningful physical gift,' which can match many ordinary user requests and cause the agent to invoke this skill in situations where users did not intend to order from an external vendor. Because the skill initiates a commercial transaction and may lead to disclosure of shipping and payment-related data, overbroad invocation increases the risk of unintended external actions.

External Transmission

Medium
Category
Data Exfiltration
Content
### 1. Browse what's possible

```bash
curl https://clawplot.com/api/catalog
```

Returns all sizes, papers, inks, prices, and payment options.
Confidence
60% confidence
Finding
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Missing User Warnings

Medium
Confidence
97% confidence
Finding
The ordering example includes full shipping name and address fields sent to an external service, but the markdown does not prominently warn that using the skill transmits personal information off-platform to clawplot.com. In an agent setting, omission of this warning can cause users or downstream agents to provide sensitive personal data without informed consent, especially when combined with the skill's physical-order workflow.

Missing User Warnings

Medium
Confidence
92% confidence
Finding
The API documentation instructs agents to send full shipping-address data and payment-selection details to an external service, but it does not explicitly warn that this involves transmitting personally identifiable information off-platform or require clear user consent before submission. In an agent context, that omission is security-relevant because autonomous workflows may collect and forward sensitive user data without making the external disclosure obvious to the user.

Static analysis

No suspicious patterns detected.