ClawHub
Back to skill

Security audit

创意总监免费版

Security checks for vulnerabilities and agentic risk

Overview

This creative brainstorming skill appears non-malicious, but it asks for shell and write capabilities that do not fit its stated Markdown-only purpose.

Review this skill before installing. Its creative-advice behavior is understandable, but the requested exec/write authority is broader than the documented need. Prefer a version that removes shell/write access or clearly limits any saving and diagnostics to user-approved actions.

SkillSpector

Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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 (4)

Description-Behavior Mismatch

Medium
Confidence
92% confidence
Finding
The skill is presented as a lightweight creative-consultation assistant, but the body introduces generic create/modify/delete/export/save operation modes that are not necessary for brainstorming. This mismatch expands the apparent authority of the skill and can mislead an agent into granting broader operational behavior than users would reasonably expect, increasing the chance of unintended file or workflow actions.

Intent-Code Divergence

High
Confidence
97% confidence
Finding
The file claims the skill is pure Markdown and does not require exec or external tooling, yet elsewhere it declares exec/write tools and includes command/network procedures. This contradiction is dangerous because it obscures the real privilege level of the skill, making risky capabilities appear harmless and undermining informed review and least-privilege enforcement.

Context-Inappropriate Capability

High
Confidence
98% confidence
Finding
Command execution capability is not justified for a creative brainstorming and evaluation skill. In this context, granting exec creates unnecessary exposure to shell command execution, local environment inspection, and possible abuse through prompt-induced tool use, far beyond what is needed to generate creative suggestions.

Context-Inappropriate Capability

Medium
Confidence
93% confidence
Finding
Write access is unnecessary for a skill whose stated function is to provide advice, brainstorming, and structured textual output. Unneeded write capability can let the agent alter local files, save misleading outputs, or modify project state in ways the user did not request, increasing integrity risk.

Static analysis

No suspicious patterns detected.