ClawHub
Back to skill

Security audit

HiFi Review

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed HiFi review helper that gathers public audio evidence and runs local analysis scripts, with no signs of credential access, persistence, exfiltration, or destructive behavior.

Install if you want audio-gear evaluations that use public measurements and reviews. Be aware it may activate automatically for HiFi-related prompts and will tend to answer bilingually with Chinese first; explicitly ask for single-language output if that matters.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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
92% confidence
Finding
The skill enables implicit invocation (`allow_implicit_invocation: true`) without any visible trigger constraints or narrowing conditions in this file. That can cause the agent to auto-route ordinary HiFi-related user queries into this skill unexpectedly, increasing the chance of unintended tool/skill activation, prompt-surface expansion, and misuse of the skill in contexts the user did not explicitly consent to.

Natural-Language Policy Violations

Medium
Confidence
91% confidence
Finding
The rule hard-codes bilingual output with Chinese first, regardless of the user's language preference or explicit opt-in. While not a classic security exploit, it can override user intent, cause unintended disclosure in shared contexts, and create prompt-compliance/privacy issues if the user expected a single-language response.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal