ClawHub
Back to skill

Security audit

Office Task Automator Free

Security checks for vulnerabilities and agentic risk

Overview

This office automation skill handles sensitive documents in bulk and should be reviewed because its local-only privacy claim is not clearly reconciled with its required LLM use.

Before installing, treat this as a powerful local file automation helper: use it only on explicit folders, back up originals, avoid sensitive HR/customer documents unless your agent platform's LLM data handling is acceptable, and require a preview before bulk writes or conversions.

SkillSpector

Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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 (3)

Intent-Code Divergence

Medium
Confidence
95% confidence
Finding
The skill claims that all data is processed locally and not uploaded to the cloud, yet later states that an LLM API is required and that the platform provides LLM calls. This creates a misleading security assurance that may cause users to submit sensitive office documents under false assumptions about data locality and exposure.

Vague Triggers

Medium
Confidence
87% confidence
Finding
The trigger language is overly broad, telling the agent to use this skill whenever file processing, format conversion, or content extraction is needed. In a tool-enabled environment with read and exec permissions, broad invocation criteria can cause the skill to activate on unintended tasks, increasing the chance of unnecessary file access, execution, or unsafe task routing.

Missing User Warnings

Medium
Confidence
91% confidence
Finding
The skill describes bulk generation, conversion, and output writing but does not consistently require confirmation or warn about large-scale file creation and modification. In practice, this can lead to accidental overwrites, uncontrolled output sprawl, or destructive processing of the wrong directory when the agent is allowed to execute commands.

Static analysis

No suspicious patterns detected.