ClawHub
Back to skill

Security audit

Pangolinfo Amazon Scraper: Products + Keywords + Reviews + Categories

Security checks across malware telemetry and agentic risk

Overview

The skill mostly matches its Amazon-scraping purpose, but its credential instructions conflict about whether the agent should directly read a Pangolinfo API key from the environment.

Review the credential setup before installing. Use a limited Pangolinfo API key, avoid exposing it in chat or logs, and prefer a single configured MCP credential path until the publisher removes the contradictory environment-variable guidance.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • 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
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Description-Behavior Mismatch

Medium
Confidence
94% confidence
Finding
The skill documentation gives contradictory guidance about secrets handling: it first states no API-key environment variable is needed, then later instructs the agent to read `PANGOLINFO_API_KEY` directly from the environment. In an agent setting, contradictory secret-handling instructions can cause unsafe implementation choices, accidental secret access, or user/operator confusion about where credentials are sourced and exposed.

Intent-Code Divergence

Medium
Confidence
95% confidence
Finding
This duplicate inconsistency is security-relevant because it mixes two trust boundaries: client-injected MCP credentials versus direct environment-variable access by the skill. That ambiguity can lead implementers to grant broader secret access than necessary or to mishandle credential configuration, violating least-privilege expectations.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.