ClawHub
Back to skill

Security audit

Kunrumary Ops Kit

Security checks across malware telemetry and agentic risk

Overview

The skill is a markdown-only healthcare marketing content generator, but it reuses detailed patient-case medical information for promotional materials without enough consent or de-identification safeguards.

Review this skill carefully before installing. It appears non-executable and not malware-like, but it should only be used with properly authorized, de-identified, and compliance-reviewed patient stories, and users should avoid relying on it for personal medical advice or unsupervised medical advertising.

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 (2)

Vague Triggers

Medium
Confidence
94% confidence
Finding
The skill advertises very broad trigger phrases such as '写保胎文案' and '保胎营销资料' without meaningful exclusion rules or disambiguation, so it may activate in loosely related medical or marketing conversations where the user did not intend to invoke this specialized skill. In a healthcare-marketing context, unintended activation is more dangerous because it can inject branded, persuasive medical content into sensitive discussions, increasing the risk of inappropriate medical marketing, privacy-adjacent mishandling, or misleading health-related outputs.

Missing User Warnings

Medium
Confidence
92% confidence
Finding
The template explicitly instructs generating marketing content from patient cases and includes fields such as age, medical history, treatment course, pregnancy milestones, delivery details, and baby characteristics. Even with a note to use pseudonyms, this combination of quasi-identifiers can enable re-identification and turns sensitive health data into promotional material without an explicit requirement for documented patient consent, data minimization, or user-facing privacy safeguards.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.