多平台翻译工具

• Outcome: Review
• Latest audit: May 28, 2026, 4:45 AM
• Version:
  1.0.0

Overview

The skill is a translator, but it misleadingly says text stays local while its instructions submit text to external translation and chat services.

Review before installing. Treat anything translated with this skill as being sent to external providers and possibly associated with logged-in Doubao or Tencent Yuanbao accounts. Do not use it for secrets, credentials, regulated personal data, confidential contracts, or sensitive business text unless those provider data flows are acceptable.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access

Findings (4)

Intent-Code Divergence

High

• Confidence: 99% confidence
• Finding: The skill explicitly sends user-provided text to external translation and AI chat services, yet the documentation claims the content stays local and is not uploaded to third-party servers. This is dangerous because users may paste sensitive business, legal, or personal data under a false privacy assumption, causing unintentional disclosure to multiple external providers.

Vague Triggers

Medium

• Confidence: 89% confidence
• Finding: The trigger set includes generic phrases related to translation that can plausibly appear in ordinary conversation, increasing the chance of accidental invocation. While this is not a direct code-execution issue, unintended activation can cause user text to be sent to external services without deliberate intent.

Missing User Warnings

High

• Confidence: 98% confidence
• Finding: The skill description and notes do not accurately warn that content will be transmitted to third-party translation websites and AI chat systems. In this context, the omission is security-relevant because the skill handles arbitrary user text, which may include confidential or regulated information.

Ssd 3

Medium

• Confidence: 99% confidence
• Finding: The instruction stating that translated content remains local directly contradicts the operational steps that submit text to external web services and authenticated chat platforms. This misleading assurance can materially increase exposure because users may trust the skill with sensitive content they would otherwise withhold.

VirusTotal

VirusTotal findings are pending for this skill version.