Search Engine

Security checks across malware telemetry and agentic risk

Overview

This is an instruction-only search-engine planning skill whose main consideration is disclosed local memory for preferences and project notes.

Before installing, decide whether you want this skill to remember search-engine project context across sessions or stay session-only. Avoid storing secrets, credentials, legal identifiers, or sensitive business details in ~/search-engine/ unless you explicitly choose to.

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 (2)

Vague Triggers

Medium

Confidence: 91% confidence
Finding: The setup instructs the skill to activate on broad concepts like 'search quality, indexing, or relevance issues' and even proactively during architecture discussions. That ambiguity can cause the skill to engage in contexts the user did not explicitly request, increasing the chance of unintended data exposure, confusion about active behavior, and unauthorized persistence of context or recommendations.

Missing User Warnings

Medium

Confidence: 88% confidence
Finding: The skill directs storage of activation preferences in memory for future sessions and later describes maintaining internal notes, but it does not clearly notify the user at the point of collection or obtain explicit consent for persistence. In a design/architecture assistant context, those notes may include project constraints, failure patterns, and workflow preferences that can be sensitive business information even if they are not secrets.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal