ClawHub
Back to skill

Security audit

QA 测试技能集

Security checks for vulnerabilities and agentic risk

Overview

This appears to be a legitimate QA test-case workflow, but it needs Review because it can auto-trigger broadly and then read local requirement trees or fetch URLs without a clear consent step.

Install only if you are comfortable with the agent reading requirement documents and linked subdocuments you provide. Use it in a limited project workspace, review links and referenced paths before running, and consider narrowing triggers or requiring confirmation before URL fetches, directory traversal, or Bash use.

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 (2)

Vague Triggers

High
Confidence
95% confidence
Finding
The activation phrases are broad enough to match ordinary requests such as '帮我测试' or '开始测试', which can cause the skill to trigger outside the user's precise intent. Because this skill can orchestrate many sub-skills and use tools like WebFetch and Bash, unintended activation can lead to unnecessary external access, excessive processing, or confusing workflow execution without clear user consent.

Missing User Warnings

Medium
Confidence
93% confidence
Finding
The skill explicitly instructs the agent to fetch and analyze URLs and to read referenced files and directories, but it provides no user-facing notice, consent gate, or origin restrictions. In a document-processing workflow, this can lead to unintended retrieval of external content or linked local materials, increasing the risk of data exfiltration, SSRF-like fetches, or over-collection of sensitive project files beyond what the user expected.

Static analysis

No suspicious patterns detected.