ClawHub

Amazon Alexa Qa

Security checks across malware telemetry and agentic risk

Overview

This skill automates Amazon shopping-assistant questions as advertised, but it also encourages scaled stealth sessions to work around rate limits.

Install only if you are comfortable automating a logged-in Amazon session and saving shopping-assistant outputs locally. Avoid using the stealth multi-session guidance, keep batches small and user-directed, review questions before submission, and delete saved result files when they are no longer needed.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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 (3)

Context-Inappropriate Capability

Medium
Confidence
96% confidence
Finding
The skill explicitly advises scaling across multiple stealth browser sessions and frames rate limits as applying per session, which facilitates evasion of platform throttling and higher-volume automated extraction. In the context of Amazon shopping AI interactions, this goes beyond ordinary single-user automation and materially increases the risk of abusive scraping, policy violations, and account or IP enforcement.

Missing User Warnings

Low
Confidence
78% confidence
Finding
The batch example writes collected Alexa/Amazon response data to a local JSON file without warning the user that collected content will persist on disk. Even if the data is not highly sensitive by default, persistent storage can expose shopping queries, recommendations, or account-contextual outputs to other local users, later processes, or unintended retention.

Missing User Warnings

Medium
Confidence
92% confidence
Finding
The script generates JavaScript that programmatically fills a shopping-assistant text area and clicks the submit button without any runtime user confirmation or visible warning. In this skill’s context, that means arbitrary prompts can be submitted to Amazon’s Alexa/Rufus interface at scale, which increases the risk of unintended actions, spammy automation, or policy-violating data collection if upstream inputs are untrusted.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal