Amsterdam

• Outcome: Pass
• Latest audit: May 28, 2026, 1:50 AM
• Version:
  1.0.1

Overview

This is a markdown-only Amsterdam guide with no code, install hooks, credentials, persistence, or privileged access, though some high-stakes advice should be verified before use.

Installers should treat this as a reference guide, not a source of legal, tax, medical, housing, or emergency advice. Verify visa, tax, healthcare, housing, safety, and cost details with official Dutch sources or qualified professionals, especially because the guide contains dated 2025-2026 figures.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• 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 (5)

Natural-Language Policy Violations

Medium

• Confidence: 86% confidence
• Finding: This section frames broad cultural traits and communication norms as if they generally apply to all Dutch people, with limited qualification or acknowledgment of variation. In a travel/culture guidance skill, that can reinforce stereotypes, mislead users in social interactions, and present a single cultural lens as universal, even if the content appears non-malicious and informational.

Missing User Warnings

Medium

• Confidence: 91% confidence
• Finding: The file lists a suicide prevention resource under mental health resources but does not immediately state that anyone in imminent danger or experiencing an acute crisis should call emergency services or seek urgent in-person help. In a healthcare guidance skill, this omission can delay lifesaving escalation because users may treat the listed resource as sufficient even during an emergency.

Natural-Language Policy Violations

Medium

• Confidence: 89% confidence
• Finding: The content recommends neighborhoods for users based on protected characteristics such as religion and ethnicity-like community affiliation ('Jewish community members', 'Japanese community members') without first establishing that the user explicitly asked for identity-based matching. In a housing context, this can facilitate steering or discriminatory personalization, which is especially sensitive because the file is directly advising where people should live.

Natural-Language Policy Violations

Medium

• Confidence: 81% confidence
• Finding: The phrase "Language barriers in some shops" presents a language-related limitation as a neighborhood drawback without offering user context, alternatives, or inclusive framing. In a relocation guidance skill, this can reinforce exclusionary bias and steer users away from diverse communities based on language assumptions rather than actionable information.

Missing User Warnings

Medium

• Confidence: 94% confidence
• Finding: This is a true issue because the document gives detailed visa, immigration, and tax guidance with concrete thresholds, fees, timelines, and eligibility claims that users may rely on for legal or financial decisions. Although it cites official sources at the end and briefly notes that some thresholds adjust annually, it does not prominently warn that immigration and tax rules change frequently, can depend on individual circumstances, and must be verified with official authorities or qualified legal/tax advisors before action.

VirusTotal

66/66 vendors flagged this skill as clean.