ClawHub
Back to skill

Security audit

文档助手工具

Security checks across malware telemetry and agentic risk

Overview

This skill is mostly a documentation helper, but it asks for command/write capability and includes an unclear script-running instruction that is not well scoped or explained.

Review this skill carefully before installing. It may be acceptable if you only need documentation navigation, but avoid granting command execution or write permissions unless the publisher provides the referenced script, explains exactly what it does, and constrains when it can run.

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)

Intent-Code Divergence

Medium
Confidence
92% confidence
Finding
The skill is presented primarily as a documentation/navigation assistant, but a usage example instructs the agent to run an unspecified script from the skill directory. That creates a hidden execution path with unclear behavior, making it easy for a user to trigger arbitrary local code execution under the trusted framing of harmless documentation help.

Description-Behavior Mismatch

Medium
Confidence
88% confidence
Finding
The skill’s stated purpose is narrow documentation assistance, but its description and declared tools expand into broad file handling, automation, writing, and command execution. This capability mismatch increases the chance that users or agents will grant or exercise privileges beyond what is reasonably necessary, enabling misuse or unexpected side effects.

Context-Inappropriate Capability

Medium
Confidence
90% confidence
Finding
A documentation expert skill declaring exec capability without a clear operational need violates least-privilege expectations. In an agent environment, command execution materially raises risk because even benign-seeming prompts can be transformed into filesystem access, script launch, or shell-based actions outside the user’s expected intent.

Missing User Warnings

Medium
Confidence
91% confidence
Finding
The documentation references command or script execution but does not clearly warn users about the security implications of running local code. This omission is dangerous because it normalizes execution inside a seemingly low-risk documentation skill, reducing scrutiny and increasing the likelihood of unsafe approval by users or orchestrators.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.