ClawHub
Back to skill

Security audit

ZeeLin Auto-PPT

Security checks across malware telemetry and agentic risk

Overview

This skill does what it advertises, but it automates a logged-in Google NotebookLM session and uploads content without a clear confirmation step.

Install only if you are comfortable letting an agent control NotebookLM in your logged-in Google account, upload the presentation content to Google, create notebook artifacts there, and save the result on your Desktop. Avoid confidential, regulated, or proprietary material unless you have reviewed and approved the upload.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (4)

Context-Inappropriate Capability

Medium
Confidence
97% confidence
Finding
The skill explicitly instructs the agent to use unrestricted shell execution as the primary control path, including passing large, model-generated content into shell commands and invoking a local script. This expands the skill from a constrained presentation workflow into arbitrary command execution, creating command-injection and unintended local-action risk if user content, filenames, or future script changes are not safely handled.

Context-Inappropriate Capability

Low
Confidence
89% confidence
Finding
The fallback section exposes direct browser-control commands that go beyond a tightly scoped skill contract and gives the agent a reusable command vocabulary for arbitrary navigation, clicking, typing, and downloading. Even if intended for NotebookLM recovery, this broadens the operational surface and makes misuse or accidental actions on other pages more likely.

Missing User Warnings

Low
Confidence
84% confidence
Finding
The README states that the skill will automatically generate and place a PDF at a fixed Desktop path, but it does not clearly warn users that a local file write will occur. Silent or insufficiently disclosed filesystem output can surprise users, overwrite expected files, or leak sensitive presentation content into an easily accessible location such as the desktop.

Missing User Warnings

Medium
Confidence
94% confidence
Finding
The script takes arbitrary user-supplied content and sends it to notebooklm.google.com via browser automation without any explicit privacy notice, consent check, or data-sensitivity warning. This is risky because users may provide confidential or regulated information, not realizing it will be transmitted to a third-party cloud service and potentially retained or processed under that provider's policies.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.