Security audit

Nano Banana 2 文生图

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed text-to-image helper that sends prompts to a fixed AI Hive API and saves generated images locally.

Before installing, understand that prompts and the AI Hive API key are used with AI Hive, and generated images are saved locally. Use an API key with appropriate account limits and avoid placing sensitive information in prompts.

SkillSpector

Vulnerability Patterns

MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access

Findings (2)

Lp3

Medium

Category: MCP Least Privilege
Confidence: 88% confidence
Finding: The skill advertises executable behavior that uses shell commands, local file writes, environment access, and outbound network requests, yet it declares no permissions. This creates a transparency and consent problem: a user or host agent may invoke it without understanding that it can store credentials locally, contact remote services, and write downloaded outputs to disk.

Tp4

High

Category: MCP Tool Poisoning
Confidence: 92% confidence
Finding: The stated purpose focuses on text-to-image generation without reference images, but the skill also performs credential setup, task polling, and local file downloads. That mismatch is risky because users may authorize what appears to be a simple generation skill without realizing it persists secrets and writes artifacts locally, increasing the chance of unintended data exposure or unsafe execution in constrained environments.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.