Security audit

Multi Search Engine

Security checks across malware telemetry and agentic risk

Overview

This search skill appears functional and purpose-aligned, but its privacy notice understates that user queries are sent to third-party search engines.

Install only if you are comfortable with search terms and request metadata being sent to the selected search engines. Avoid secrets, credentials, internal hostnames, confidential business topics, regulated data, or sensitive personal queries unless the publisher corrects the privacy notice and adds clearer provider routing controls.

SkillSpector

By NVIDIA

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (3)

Intent-Code Divergence

Medium

Confidence: 97% confidence
Finding: The documentation states that there is 'no external data transmission', but the skill explicitly sends user queries and HTTP requests to third-party search engines. This is a misleading privacy claim that can cause users or downstream agents to disclose sensitive queries under false assumptions about network exposure.

Natural-Language Policy Violations

Medium

Confidence: 84% confidence
Finding: The skill automatically routes queries to different search providers based on detected language without explicit user choice. This can leak user queries to jurisdictions or providers the user did not intend to use, creating privacy, compliance, and data-handling risks.

Missing User Warnings

Medium

Confidence: 88% confidence
Finding: The markdown includes executable-looking examples that send user-supplied queries to external search engines, but it does not warn that those queries may disclose sensitive research topics, internal domains, or other private data to third parties. In an agent skill context, users may copy these patterns directly into automated workflows, increasing the chance of unintended data leakage and external tracking.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.