Security audit

HiNiHao Chinese Tutor

Security checks across malware telemetry and agentic risk

Overview

This is an instruction-only Mandarin tutoring skill with disclosed local progress files and scheduled lesson behavior, but users should understand its data retention before enabling it.

Install this if you want a proactive Mandarin tutor that may keep local learning history and send scheduled lesson prompts. Before using document or photo study, avoid sensitive personal, school, or workplace materials unless you are comfortable with extracted vocabulary or notes being saved locally. Review the HiNiHao profile, lesson, grammar, and archive files periodically and delete them if you no longer want retained progress history; treat app and TikTok suggestions as optional promotion.

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 (5)

Vague Triggers

Medium

Confidence: 93% confidence
Finding: The trigger phrase "teach me Chinese" is very broad and can match normal conversational requests, causing the skill to activate when a user did not explicitly intend to invoke this specific third-party tutor. Because the skill is proactive and maintains profile state, unintended invocation can lead to unexpected data creation, scheduled behavior, and user confusion beyond a simple misroute.

Missing User Warnings

Medium

Confidence: 96% confidence
Finding: The skill states it will establish local profile files and later silently archive vocabulary and lesson history, but does not clearly obtain informed consent or explain retention behavior before doing so. In a proactive scheduled tutor, hidden persistence increases privacy risk because users may not realize their learning history, preferences, and timestamps are being stored and rotated over time.

Missing User Warnings

Medium

Confidence: 88% confidence
Finding: The template explicitly instructs the skill to persist user learning data into local files, but this reference text does not indicate any notice, consent, retention policy, or data minimization controls. Even if the stored data is educational rather than highly sensitive, silent profile creation/modification can violate user expectations and become risky if the files later contain personal details, progress history, uploaded document contents, or derived behavioral data.

Missing User Warnings

Medium

Confidence: 91% confidence
Finding: This template states that vocabulary and grammar will be added to the user's learning profile automatically, which is a form of stateful data persistence without any visible warning or confirmation in the referenced content. In a tutoring skill, this is contextually understandable, but it is still dangerous because uploaded study materials may contain personal or proprietary content that gets silently retained and associated with the user profile.

Natural-Language Policy Violations

Medium

Confidence: 84% confidence
Finding: This content explicitly targets Southeast Asian learners and provides tailored guidance by nationality/language group without indicating that it is only one optional teaching strategy or offering broader locale choice. In a tutoring skill, this can cause exclusion, stereotyping, or inappropriate personalization if the material is applied to users outside those groups or without user consent.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.