ClawHub
Back to skill

Security audit

文档解析工具（免费版）

Security checks for vulnerabilities and agentic risk

Overview

This document-parsing skill is not clearly malicious, but it asks for broad command/file/API capabilities and does not clearly bound cloud OCR handling of potentially sensitive documents.

Install only if you are comfortable reviewing each run manually. Avoid using it on confidential, legal, financial, medical, or proprietary documents unless you can force local-only processing and know exactly which commands or cloud OCR/API services will be used.

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • 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
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)

Description-Behavior Mismatch

Medium
Confidence
93% confidence
Finding
The skill is presented as a narrowly scoped document parsing/OCR tool, but the body broadens into generic create/query/export/modify/delete-style automation without clear boundaries. This scope drift can cause an agent to invoke exec/write-enabled workflows beyond user expectations, increasing the chance of unintended file operations or command execution.

Context-Inappropriate Capability

Medium
Confidence
94% confidence
Finding
The skill introduces API-key setup, external service usage, and network troubleshooting despite being described primarily as a document parsing tool. Because documents may contain sensitive content, undocumented or weakly disclosed transmission to cloud OCR/API services creates a meaningful confidentiality risk.

Intent-Code Divergence

Medium
Confidence
89% confidence
Finding
The skill markets itself as a limited free edition for single-file/core functionality, but later claims broad automation, API integration, command execution, information retrieval, and file writing as core features. This inconsistency obscures the real privilege and data-handling surface, making risky behavior easier to trigger unintentionally.

Missing User Warnings

High
Confidence
97% confidence
Finding
The documentation references external OCR/API services and API keys but does not clearly warn users that uploaded document contents may leave the local environment. In the context of document parsing, this is especially dangerous because files often contain personal, financial, legal, or proprietary information.

Static analysis

No suspicious patterns detected.