ClawHub

免费版

Security checks across malware telemetry and agentic risk

Overview

This is a legitimate data-analysis skill, but it understates privacy risk when uploaded datasets may be processed by configured AI providers.

Review before installing if you plan to analyze confidential, regulated, customer, financial, or proprietary data. Uploaded file contents, row samples, prompts, and derived reports may be sent to your configured AI provider, so use anonymized data or a provider whose privacy, retention, and compliance terms you accept.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (4)

Vague Triggers

High
Confidence
95% confidence
Finding
The trigger list is broad enough to match common, non-specific user requests such as 'find insights' or '帮我分析', which can cause the skill to activate unexpectedly for unrelated conversations involving uploaded files or general analysis. This increases the chance of unintended data handling and prompt hijacking through ambiguous invocation, especially because the skill is designed to immediately ingest and summarize uploaded content.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The document reassures users that data only flows between the local machine and the AI, but it also references configuring external model providers such as 阿里云百炼 or DeepSeek. Without an explicit privacy warning, users may incorrectly assume their files never leave their environment, leading to unintended transmission of sensitive or regulated data to third-party model services.

Vague Triggers

Medium
Confidence
84% confidence
Finding
The trigger phrases are extremely generic (for example, '帮我分析这个数据' and similar everyday requests), which increases the chance that the skill activates in contexts broader than the user intended. In a data-analysis skill, overbroad activation can cause users to send sensitive business or personal data to the configured model without a clear, deliberate invocation boundary.

Missing User Warnings

Medium
Confidence
94% confidence
Finding
The skill description emphasizes convenience and local use, but does not clearly warn up front that uploaded or pasted data will be transmitted to whatever AI provider the user configured. Because this skill is specifically designed to handle arbitrary Excel, CSV, and JSON files, the missing disclosure materially increases the risk of sensitive data exposure, regulatory issues, and user misunderstanding about where their data goes.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal