ClawHub

Product Requirement Miner

Security checks across malware telemetry and agentic risk

Overview

This appears to be a local review-analysis skill whose file reads and generated outputs fit its purpose, but users should treat the saved review files as potentially sensitive.

Install only if you are comfortable with the skill creating local files that may contain raw or derived customer review content. Use non-sensitive or redacted CSVs when possible, choose a controlled working directory, and delete generated intermediate files when you no longer need them.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • 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)

Lp3

Medium
Category
MCP Least Privilege
Confidence
84% confidence
Finding
The skill clearly instructs reading a user-supplied CSV and writing multiple local output files, but it does not declare permissions or otherwise signal those file access capabilities. That reduces transparency and makes it harder for a host or user to assess data exposure risk before execution, especially because review data may contain personal or commercially sensitive content.

Missing User Warnings

Medium
Confidence
92% confidence
Finding
The skill instructs saving raw review data to `raw_reviews.txt` without any warning, retention guidance, or redaction step. Raw product reviews can contain personal data, account details, internal business information, or customer complaints, so persisting them unprotected on disk increases privacy and data leakage risk.

Missing User Warnings

Medium
Confidence
94% confidence
Finding
The workflow persists several derivative files (`classified_reviews.json`, `filtered_data.txt`, clustering reports, and roadmaps) that may still contain sensitive or identifying review content, yet it provides no warning about on-disk persistence or downstream sharing risk. Multiple artifacts expand the attack surface because sensitive text is duplicated across files and may be retained longer than intended.

VirusTotal

67/67 vendors flagged this skill as clean.

View on VirusTotal