ClawHub
Back to skill

Security audit

Skill编写工具免费版

Security checks across malware telemetry and agentic risk

Overview

This skill mostly matches skill-authoring work, but it asks for broad write and command-execution authority with weak scoping and mixed trigger language.

Review before installing. Use it only in a workspace where creating or modifying skill files is intended, and require explicit confirmation before it runs shell commands, imports or exports content, resets configuration, or deletes anything.

SkillSpector

Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
  • 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
Findings (3)

Description-Behavior Mismatch

Medium
Confidence
86% confidence
Finding
The skill is presented as a lightweight writing/skill-authoring tool, but its documented operations broaden into generic create/query/modify/delete and import/export/save behavior. This capability drift can cause an agent to grant the skill wider authority than users expect, increasing the chance of unintended file or data manipulation when the skill is selected for benign content tasks.

Context-Inappropriate Capability

Medium
Confidence
92% confidence
Finding
The skill declares exec capability even though its stated purpose is lightweight skill writing and content creation. Exposing command execution in a skill that does not strictly require it expands the attack surface substantially, because an agent may execute shell commands, create files, or alter the environment under a misleadingly low-risk description.

Missing User Warnings

Medium
Confidence
88% confidence
Finding
The skill includes instructions for creating directories, touching files, and using executable scripts without clearly warning users about filesystem or system impact. In an agent context, omission of such warnings is dangerous because users may invoke the skill expecting harmless drafting help while the agent performs state-changing local actions.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.