ClawHub
Back to skill

Security audit

Paper Parse

Security checks across malware telemetry and agentic risk

Overview

This appears to be a normal academic-paper parsing skill, with some usability and invocation-scope cautions but no artifact-backed malicious behavior.

Install only if you want an agent to fetch or process paper PDFs and convert them to text. Prefer explicit prompts that name the PDF or URL, avoid sensitive unpublished documents unless you trust the runtime, and ensure paths/URLs are handled safely by the platform or shell wrapper.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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)

Vague Triggers

Medium
Confidence
94% confidence
Finding
The trigger condition is extremely broad: it activates for essentially any request to analyze, interpret, or summarize any academic paper. Overly broad triggers can cause inappropriate auto-selection of the skill, increasing the chance that it handles requests outside its intended scope, consumes unnecessary tools/resources, or overrides more suitable workflows. In this context, the skill also instructs downloading/parsing PDFs and URLs, which makes accidental invocation more operationally risky.

Natural-Language Policy Violations

Medium
Confidence
87% confidence
Finding
The skill description and instructions mandate Chinese output without giving the user a language choice or documenting a necessary locale restriction. This is a policy and usability weakness because it can override user preference, cause misunderstandings, and produce outputs the user cannot effectively review, especially for technical or safety-relevant content. The issue is less severe than code-execution risks, but it still degrades safe and reliable operation.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.