Back to skill

Security audit

Nano Banana Pro 海报生成

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed AI Hive image/poster generator that stores an API key only through an interactive setup flow and uploads/downloads media for user-directed generation tasks.

Install only if you are comfortable using AI Hive: the skill may open a browser for API key setup, store that key locally, upload prompts and chosen reference images to AI Hive, and download generated images to your machine. Avoid uploading private or unlicensed material.

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 (3)

Lp3

Medium
Category
MCP Least Privilege
Confidence
81% confidence
Finding
The skill invokes shell commands, installs dependencies, reads environment variables, writes local configuration, and performs network operations, yet no permissions are declared to inform or constrain those capabilities. This creates a transparency and governance gap: users may run a seemingly simple poster-generation skill without understanding that it can modify the local environment and communicate externally.

Tp4

High
Category
MCP Tool Poisoning
Confidence
92% confidence
Finding
The declared purpose is narrow image/poster generation, but the referenced workflow and underlying client support broader behaviors including API key setup, browser launching, media upload, task polling, and unrelated capabilities such as chat, video generation, model listing, and user info retrieval. This mismatch is dangerous because it widens the effective attack surface beyond user expectations, enabling credential handling and external data transfer under the cover of an art-generation skill.

Description-Behavior Mismatch

High
Confidence
92% confidence
Finding
The file contains a full generic AI Hive client with capabilities for text chat, video generation, model enumeration, user info retrieval, and arbitrary media upload, even though the advertised skill is narrowly described as Nano Banana Pro poster generation. In an agent-skill context, hidden surplus capabilities expand the attack surface and enable misuse beyond user expectations, including account probing, broader API use, and unintended data transfer.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.