ClawHub
Back to skill

Security audit

The Churning Customer

Security checks across malware telemetry and agentic risk

Overview

This is a text-only product strategy skill that simulates a hypothetical churning customer and does not request system access, credentials, persistence, or external data transfer.

Installers should understand that this skill generates plausible synthetic customer narratives, not factual churn analysis. Use it for ideation and interview planning, and validate outputs against real customer data before making product or business decisions.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (1)

Vague Triggers

Medium
Confidence
92% confidence
Finding
The skill description includes broad trigger phrases such as 'why do customers really churn' and 'what does silent churn look like for my product,' which can match ordinary product-strategy or customer-research requests rather than a narrowly scoped roleplay intent. This can cause unintended invocation, leading the agent to apply a speculative churn-simulation workflow when the user may have wanted factual analysis, potentially producing misleading or overconfident synthetic customer narratives.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.