ClawHub
Back to skill

Security audit

广告情报专业版

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed API client for ad intelligence data, with normal cautions around sending queries and an API key to a third-party service.

Install only if you intend to use the paid ad intelligence API. Treat searches, product identifiers, market filters, and API responses as data sent to a third-party service, keep ADC_INTEL_API_KEY in an environment or credential store, and review commands before running bulk export or scheduled dashboard workflows.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • 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
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (3)

Intent-Code Divergence

Medium
Confidence
89% confidence
Finding
The skill claims user input should not be concatenated into command parameters, yet its documented shell workflows interpolate variables directly into curl payloads and command arguments. In an MD+EXEC skill, this inconsistency is dangerous because an agent may map untrusted user-provided values into shell context, creating command-injection risk or malformed requests that exfiltrate data.

Vague Triggers

Medium
Confidence
85% confidence
Finding
The scope section includes a long list of vague, fragmented keywords that can match broad user requests unrelated to the intended API workflows. Overbroad invocation criteria increase the chance the skill is auto-selected in inappropriate contexts, causing unintended external API use, unnecessary credential exposure, or execution of networked workflows when the user did not explicitly request them.

Missing User Warnings

Medium
Confidence
93% confidence
Finding
The markdown includes curl examples that transmit an API key and query data to an external domain, but it does not prominently warn that user inputs and results leave the local/system boundary. In an agent skill with exec capability, unclear disclosure can mislead users into sending sensitive business queries or identifiers to a third-party service without informed consent.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.