Renxing Shangzhan

Security checks across malware telemetry and agentic risk

Overview

This is a disclosed, local marketing-analysis skill with ethically sensitive persuasion advice, but it shows no hidden execution, data theft, persistence, or privileged access.

Install only if you specifically want a persuasion-focused business and marketing analysis skill. Treat its outputs as ethically sensitive: avoid targeting vulnerable groups, deceptive scarcity, fear pressure, or coercive retention tactics, and consider narrowing the triggers so ordinary analysis requests do not activate it unexpectedly.

SkillSpector

By NVIDIA

Vulnerability Patterns

Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
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

Findings (10)

Context-Inappropriate Capability

Medium

Confidence: 97% confidence
Finding: The advisor flow goes beyond descriptive analysis and explicitly recommends how to exploit fear, greed, vanity, sunk-cost effects, belonging, and other psychological weaknesses to drive acquisition, conversion, repurchase, and virality. In this skill context, that is not neutral market education: it operationalizes manipulative persuasion tactics against targeted groups, including potentially vulnerable populations such as students, older adults, and mothers.

Intent-Code Divergence

Medium

Confidence: 91% confidence
Finding: The help text presents the tool as business-mode analysis and deconstruction, but the implemented advisor mode generates concrete execution tactics for manipulating user behavior and increasing payment/retention. That mismatch can mislead reviewers and users about the tool's actual function, making harmful capability easier to distribute or invoke under an innocuous description.

Vague Triggers

High

Confidence: 96% confidence
Finding: The trigger list contains highly generic terms like “分析/analyze,” “营销,” and “人性,” which are likely to match many unrelated user requests and cause the skill to activate outside its intended scope. In this context, the skill is specifically focused on psychological triggers, human weaknesses, and business strategy, so accidental invocation could route ordinary analysis or marketing queries into manipulative framing and advice.

Vague Triggers

Medium

Confidence: 88% confidence
Finding: The documentation repeats broad trigger conditions without clarifying what should or should not activate the skill, reinforcing ambiguous matching behavior. Because the skill’s subject matter involves exploiting human triggers and weaknesses, unclear boundaries increase the chance of inappropriate activation on benign business or analytical prompts.

Ssd 4

Medium

Confidence: 98% confidence
Finding: This section lays out a multi-step funnel—targeting fears/desires, triggering demand, inducing payment, promoting repurchase, and driving referrals—explicitly framed around exploiting human vulnerabilities. The structure increases danger because it turns psychological manipulation into a repeatable operational playbook rather than isolated commentary.

Ssd 4

Medium

Confidence: 95% confidence
Finding: The fusion feature scores combinations of human triggers by 'power' and then translates them into executable market strategies, effectively helping a user optimize manipulative influence techniques. In the context of a skill explicitly about 'human weakness' and 'business warfare,' this materially increases the likelihood of abusive marketing or coercive product design.

Overly Broad Trigger

Low

Category: Trigger Abuse
Confidence: 84% confidence
Finding: The single-word trigger “人性” is extremely broad and can appear in many harmless discussions about psychology, literature, philosophy, or social behavior. Since this skill is oriented toward business warfare and psychological inducement, accidental activation could expose users to manipulative-analysis content they did not request.

Overly Broad Trigger

Low

Category: Trigger Abuse
Confidence: 85% confidence
Finding: The trigger “诱因” is a common term that may appear in routine conversations about causes, motivations, or psychology, making unintended activation likely. In a skill centered on combining psychological triggers for business influence, this broad trigger raises misuse and misrouting risk.

Overly Broad Trigger

Low

Category: Trigger Abuse
Confidence: 98% confidence
Finding: The trigger “分析” is extremely generic and is likely to match a vast number of ordinary requests across nearly all domains. Given that this skill frames outputs around human weakness, persuasion, and competitive business tactics, unintended activation could systematically steer unrelated analytical queries into psychologically manipulative or ethically risky guidance.

Overly Broad Trigger

Low

Category: Trigger Abuse
Confidence: 90% confidence
Finding: The single-word trigger “营销” is broad and may capture many legitimate marketing questions that are not seeking this skill’s specialized focus on human weaknesses and psychological pressure points. That mismatch can increase the chance of overreaching or manipulative advice being injected into otherwise standard marketing discussions.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal