ClawHub
Back to skill

Security audit

paper-reading

Security checks across malware telemetry and agentic risk

Overview

This skill is a straightforward paper-reading helper that downloads or reads PDFs and saves notes where the user asks, with ordinary file overwrite caution.

Before installing, understand that this skill can contact arXiv, install or use pdfplumber, read PDFs you provide, and write PDFs, extracted text, and notes to paths you specify. Choose non-sensitive inputs when possible and avoid pointing outputs at important existing files.

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 (2)

Lp3

Medium
Category
MCP Least Privilege
Confidence
94% confidence
Finding
The skill instructs the agent to fetch PDFs from arXiv or search by paper title and to save both downloaded PDFs and generated notes to user-specified absolute paths, which clearly exercises network access and file writes without any declared permission model. This is dangerous because it can lead to unexpected outbound requests and arbitrary file creation/overwrite behavior that users or the platform may not have explicitly authorized.

Missing User Warnings

Low
Confidence
92% confidence
Finding
The script downloads remote content and writes it to a user-controlled output path, or to an auto-generated filename in the current working directory, without any confirmation or overwrite protection. In an agent context, this can unexpectedly clobber existing local files or create files in sensitive locations if the caller supplies a dangerous path, causing data loss or unsafe filesystem side effects.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.