Gemini Watermark

Security checks across malware telemetry and agentic risk

Overview

This is a local image-editing skill that openly removes visible Gemini watermarks, which has provenance caveats but no hidden system, credential, network, or persistence behavior.

Install only if you are comfortable with a tool whose main purpose is removing visible AI watermarks. Use it only on images you are authorized to edit, keep originals, and check whether removing attribution or provenance markers conflicts with platform rules or disclosure expectations. Install dependencies in a virtual environment from trusted package sources.

SkillSpector

By NVIDIA

Vulnerability Patterns

Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
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 (1)

Natural-Language Policy Violations

High

Confidence: 99% confidence
Finding: This skill is explicitly designed to remove a visible Gemini watermark from generated images, which directly circumvents provenance and attribution safeguards. The surrounding context increases risk because the documentation operationalizes the removal at scale (including batch mode, force flags, and custom alpha maps), making misuse straightforward for concealment, policy evasion, and deceptive redistribution.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal