ClawHub
Back to skill

Security audit

Landing Optimizer

Security checks across malware telemetry and agentic risk

Overview

This skill is a marketing landing-page optimizer that stores campaign notes in declared memory locations, with no executable code or hidden high-impact behavior found.

Before installing, be aware that this skill is designed to remember landing-page plans and selected campaign facts for later marketing workflows. Avoid using it with confidential promo codes, unpublished campaign URLs, or sensitive performance data unless you are comfortable with those details being saved in the agent memory files or you instruct the agent not to save them.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • 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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (1)

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill explicitly writes campaign details, URLs, conversion rates, promo codes, and creator identifiers to persistent memory files and promotes some of them into a hot cache, but it provides no user-facing notice or consent checkpoint. In agent environments, silent persistence can expose sensitive business data across later sessions, broaden access beyond the original task, and violate least-knowledge and data-minimization expectations.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.