ClawHub
Back to skill

Security audit

亚马逊-爆款趋势挖掘师 | 青虎AI

Security checks for vulnerabilities and agentic risk

Overview

This skill is mostly coherent for Amazon product research, but it tells the agent to create local export files for larger result sets without asking the user first.

Install only if you are comfortable giving the skill access to your Qinghu API token and having larger product-research results exported to local files automatically. Avoid using it for confidential searches unless you control where exports and cached files are stored and cleaned up.

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
  • 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
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (1)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The skill explicitly instructs the agent to automatically export a table file whenever the result set has 10 or more records, even if the user did not request export, and without a clear user-facing warning that a local file may be created. This can cause unintended persistence of potentially sensitive business query results on the host filesystem and may also trigger downstream file-handling behavior without informed consent.

Content

No source excerpt is available for this finding.