nanobanana2

Security checks across malware telemetry and agentic risk

Overview

This skill is a straightforward Google Gemini image generation/editing wrapper with disclosed API-key use and third-party prompt/image processing.

Install only if you are comfortable sending prompts and any input images you choose to edit to Google's Gemini API. Prefer using GEMINI_API_KEY or a platform secret store instead of pasting API keys into chat or command arguments, and avoid confidential or regulated images unless your Google API terms and privacy requirements allow it.

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 (1)

Lp3

Medium

Category: MCP Least Privilege
Confidence: 89% confidence
Finding: The skill documentation instructs use of an environment variable (`GEMINI_API_KEY`) and an explicit `--api-key` parameter, but the skill declares no permissions or capability metadata for environment access. That mismatch can cause secret-handling behavior to be hidden from the permission model and makes it easier for agents to access or request sensitive credentials without transparent review.

VirusTotal

67/67 vendors flagged this skill as clean.

View on VirusTotal