Expanso language-detect

• Outcome: Review
• Latest audit: May 28, 2026, 11:35 AM
• Version:
  1.0.0

Overview

The skill appears to detect language as claimed, but its server mode can expose an unauthenticated endpoint that sends submitted text to OpenAI using the user's API key.

Use CLI mode for controlled one-off checks where possible. If running the MCP/HTTP mode, bind it to localhost or protect it with firewall/auth controls, use a restricted or low-quota OpenAI key, and avoid sending confidential, regulated, or secret-containing text unless OpenAI processing is approved.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (1)

Missing User Warnings

High

• Confidence: 96% confidence
• Finding: The pipeline forwards user-supplied text directly to the OpenAI API, which transfers potentially sensitive content to a third-party service without any visible disclosure, consent mechanism, minimization, or redaction. In a language-detection skill, full raw text is often unnecessary or higher-risk than needed, so this design can expose secrets, personal data, or regulated content if users submit them.

VirusTotal

66/66 vendors flagged this skill as clean.