ClawHub
Back to skill

Security audit

报表配置基础版

Security checks across malware telemetry and agentic risk

Overview

This report-generation skill is mostly coherent, but it requests shell execution and leaves remote-data behavior under-scoped for a basic reporting workflow.

Review this before installing if you will use it on sensitive reports. The report export behavior is expected, but shell execution should be removed or tightly limited, and remote data access should be treated as off unless you explicitly intend it and understand what data may leave the local environment.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (4)

Context-Inappropriate Capability

Medium
Confidence
90% confidence
Finding
The skill requests shell execution capability even though its stated function is basic single-source report parsing, aggregation, and export. This creates unnecessary command-execution attack surface: if later prompts or implementations map user-controlled inputs into exec calls, the skill could enable arbitrary command execution or unsafe local actions unrelated to reporting.

Intent-Code Divergence

Low
Confidence
70% confidence
Finding
The documentation inconsistently states that no extra API key is needed while also referencing remote data source/network errors and LLM API dependency. This ambiguity can mislead users about when data may leave the local environment, increasing the risk of unintentional disclosure of sensitive report inputs to remote services.

Missing User Warnings

Medium
Confidence
87% confidence
Finding
The skill declares file write and command execution capabilities but does not prominently warn users that it may modify local files or invoke shell actions. For a reporting skill, these powerful capabilities are contextually broader than necessary and could surprise users, especially if operating on sensitive directories or if combined with prompt-controlled parameters.

Missing User Warnings

Medium
Confidence
80% confidence
Finding
The skill mentions remote data source handling and network-related failures without a privacy warning that user data may be fetched from or transmitted over a network. In a reporting workflow, source files may contain sensitive business data, so undisclosed remote access meaningfully increases confidentiality risk.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.