Security audit

ADC研发专利与交易监控（ADC R&D, Patent & Deal Monitoring）

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed ADC monitoring/report workflow that queries domain and web sources and generates an HTML brief, with no evidence of hidden persistence, credential access, or destructive behavior.

Install only if you want broad ADC monitoring reports that query PatSnap/Zhihuiya-style MCP data sources plus web search and produce downloadable HTML briefs. Prefer invoking it explicitly with @skill:adc-landscape-monitoring when you want the full workflow.

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: 95% confidence
Finding: The trigger phrase "ADC最新进展" is broad enough to match normal user intent for general domain information rather than an explicit request to invoke this skill. That increases the chance of unintended activation, causing the agent to run a large multi-step workflow, collect external data, and generate reports when the user may have only wanted a simple answer.

Vague Triggers

Medium

Confidence: 94% confidence
Finding: The example phrase "过去一个月ADC进展" reads like an ordinary analytical query, not a deliberate tool invocation. In this skill's context, accidental matching is more dangerous because activation launches broad parallel searches, external web queries, and report generation across multiple data sources, potentially wasting resources and surprising users.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.