ClawHub
Back to skill

Security audit

卖货文案匠

Security checks across malware telemetry and agentic risk

Overview

This sales-copy skill is mostly documentation-only, but it asks for command execution and file access that are not clearly needed for generating marketing text.

Install only if you are comfortable granting a marketing-copy skill read and command-execution capability. Prefer running it in a constrained agent environment, deny exec unless you specifically need it, and do not expose sensitive environment variables or API keys to this skill without reviewing its behavior.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • 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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Findings (4)

Context-Inappropriate Capability

Medium
Confidence
91% confidence
Finding
The skill’s stated purpose is sales-copy generation, yet it requests exec and read capabilities and describes file operations and command invocation unrelated to that narrow function. This unnecessary expansion of capability increases attack surface and can enable command execution or filesystem access if an agent blindly grants the declared tools.

Intent-Code Divergence

Medium
Confidence
88% confidence
Finding
The documentation is internally inconsistent: it claims the skill is 'pure LLM-driven' with no extra dependencies while also requiring exec for file writes and command calls. Misrepresenting operational behavior can cause users or agent frameworks to underestimate risk and approve dangerous permissions they would otherwise deny.

Intent-Code Divergence

Medium
Confidence
90% confidence
Finding
Later sections broaden the skill from copywriting into API connections, file processing, multi-interface aggregation, and command execution, none of which are established in the core workflow. This scope creep is dangerous because it normalizes broader automation powers under a benign marketing-skill label, making overprivileged deployment more likely.

Missing User Warnings

Medium
Confidence
86% confidence
Finding
The skill declares exec for file writes and command execution without a prominent warning that it may perform system-impacting actions. In agent environments, this can mislead operators into enabling a high-risk tool for a task that appears to require only text generation.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.