Security audit

HUMAN 3.0 Life Planner

Security checks across malware telemetry and agentic risk

Overview

This is a text-only Chinese life-assessment skill, but its metadata declares unrelated crypto and purchase capabilities users should review before installing.

Install only if you want a Chinese-language structured personal-development assessment and are comfortable sharing sensitive personal details in conversation. Before enabling it, review platform permissions carefully; the declared crypto and purchase capabilities do not fit the skill content and should not be granted without a clear reason.

SkillSpector

By NVIDIA

Vulnerability Patterns

Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access

Findings (5)

Vague Triggers

Medium

Confidence: 93% confidence
Finding: The activation text includes many broad phrases such as general life guidance requests plus a catch-all condition for any comprehensive life evaluation using this framework. That can cause unintended invocation during ordinary conversations, exposing users to a highly directive assessment flow and sensitive probing without clear consent. In this context, the skill is especially risky because it performs multi-step psychological-style interviewing and produces authoritative-seeming development judgments.

Natural-Language Policy Violations

Medium

Confidence: 90% confidence
Finding: Mandating Chinese for all interactions without checking user preference can exclude or confuse users and may cause them to misunderstand sensitive guidance. In a life-assessment skill, misunderstanding questions, warnings, or recommendations can materially distort the assessment and lead to harmful or inappropriate advice.

Natural-Language Policy Violations

Medium

Confidence: 91% confidence
Finding: The required opening script forces Chinese delivery before establishing that the user understands or wants Chinese. Because this script initiates a psychologically intensive assessment and frames the interaction as direct truth-telling, a language mismatch increases the chance of invalid consent, misinterpretation, and distress.

Natural-Language Policy Violations

Medium

Confidence: 89% confidence
Finding: Requiring the final report to be only in Chinese can make conclusions and action steps inaccessible to some users, especially if the interaction was triggered unintentionally or the user is multilingual but prefers another language. In this skill, the report contains development judgments and risk-related recommendations, so comprehension is important for safety and informed use.

Natural-Language Policy Violations

Medium

Confidence: 93% confidence
Finding: The file hard-codes Chinese-language prompts and repeatedly instructs the assessor to conduct the interview in Chinese, without indicating that the user should be asked for language preference first. This can create exclusion, misunderstanding, or unsafe miscommunication for users who triggered the skill via English phrases or who are not comfortable discussing sensitive life-assessment topics in Chinese.

VirusTotal

50/50 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.