ClawHub
Back to skill

Security audit

测试2123

Security checks for vulnerabilities and agentic risk

Overview

The skill appears to be a planning aid, but it overreaches by broadly activating for creative work and directing persistent repository changes without clear user approval.

Install only if you want a planning skill that may create planning files. Before use, require the agent to ask before writing files or making git commits, and use your preferred language so you can review the plan and any repository changes clearly.

SkillSpector

Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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 (3)

Vague Triggers

Medium
Confidence
92% confidence
Finding
The skill declares itself mandatory for essentially any creative work, which can cause overbroad interception of user tasks and unnecessary collection of project context before work begins. In an agent setting, broad mandatory activation increases the chance of workflow hijacking, user-friction, and inappropriate invocation on tasks where this skill is irrelevant or harmful.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill instructs the agent to write a file under docs/plans and commit it to git as part of normal flow, but it does not require explicit user approval before making repository changes. This is dangerous because it can lead to unintended persistence, unauthorized commits, and modification of the user's workspace from what should be a conversational brainstorming step.

Natural-Language Policy Violations

High
Confidence
89% confidence
Finding
The skill content is entirely in Chinese and does not offer a language choice or document a justified locale constraint, which can override user preference and reduce transparency about what the agent is doing. While not typically a direct security exploit, forced language can impair informed consent, reviewability, and safe human oversight of actions such as file creation or workflow changes.

Static analysis

No suspicious patterns detected.