ClawHub
Back to skill

Security audit

Walmart Keyword Search

Security checks across malware telemetry and agentic risk

Overview

The skill mostly does what it says, but it includes explicit guidance for stealth browser sessions to work around scraping limits, so users should review it before installing.

Install only if you are comfortable with a Walmart scraping helper that may be used for bulk collection. Remove or ignore the stealth-session and fingerprint-splitting guidance, keep runs modest, respect Walmart access limits and terms, and review the local memory file behavior if you do not want cross-task notes retained.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)

Context-Inappropriate Capability

Medium
Confidence
95% confidence
Finding
The skill explicitly advises anti-scraping evasion tactics such as using multiple stealth browser sessions and distributing work across independent fingerprints. That goes beyond ordinary extraction instructions and materially increases the agent's ability to avoid detection and rate-limiting controls imposed by the target site.

Vague Triggers

Medium
Confidence
79% confidence
Finding
The trigger phrases are very broad and include common shopping and research requests, which increases the chance the skill will auto-activate in contexts where the user did not specifically request scraping. Because the skill performs extraction from a third-party site, overbroad invocation expands the operational surface and can cause unintended or policy-violating data collection.

Missing User Warnings

Medium
Confidence
91% confidence
Finding
The skill instructs the agent to read from and append to a local memory file without requiring user notice or consent. Silent filesystem writes create persistence, may leak operational history across tasks, and establish behavior outside the user's expected scope for a search-scraping skill.

Ssd 4

Medium
Confidence
94% confidence
Finding
The batching guidance is not purely operational; it instructs the agent to avoid parallelism in one browser while increasing throughput through multiple stealth sessions with independent fingerprints. This is a concrete anti-detection strategy that helps scale scraping while reducing the likelihood of enforcement by the target platform.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.