ClawHub
Back to skill

Security audit

ARTi 机会雷达

Security checks across malware telemetry and agentic risk

Overview

This skill is a coherent finance-news assistant that stores a local watchlist profile with disclosure and confirmation, while avoiding credentials, trading actions, and hidden data transfer.

Before installing, be aware that the skill may save your watchlist, holdings names, themes, locale, timezone, and source choices in a hidden profile file in the current project. Avoid entering broker credentials or account values. If ambiguous trigger phrases cause unwanted activation, invoke it with explicit phrases like "机会雷达" or "股票早报" and review the profile before confirming changes.

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 (4)

Vague Triggers

Low
Confidence
90% confidence
Finding
The default prompt uses a broad invocation phrase that can trigger the skill in loosely related contexts, increasing the chance of over-invocation or accidental routing. In a finance/news skill, this can expose portfolio-related context or cause unsolicited market-analysis behavior when the user did not clearly request it.

Natural-Language Policy Violations

Low
Confidence
78% confidence
Finding
Mixing Chinese display metadata with an English default prompt without an explicit language-selection mechanism can cause user confusion about what the skill will do and when it should be invoked. That ambiguity can lead to mistaken consent or unintended activation, especially for users who rely on the displayed language to assess relevance and trust.

Vague Triggers

High
Confidence
95% confidence
Finding
Several triggers such as "早报", "晚报", "今日新闻", and "找机会" are generic enough to match broad, unrelated user requests. In a finance skill, this can cause unintended invocation on ambiguous prompts, leading the agent to access portfolio/watchlist context or steer the conversation into investment workflows the user did not explicitly request.

Natural-Language Policy Violations

Medium
Confidence
88% confidence
Finding
The schema sets `locale` to `zh-CN` by default without indicating that this value must come from explicit user choice or informed consent. In a skill that stores a persistent user profile, hard-coding locale can incorrectly infer nationality/language preferences and may drive downstream personalization, source selection, or reporting behavior in ways the user did not request.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.