ClawHub
Back to skill

Security audit

Nano Banana 2 角色一致性图片

Security checks across malware telemetry and agentic risk

Overview

The skill is a purpose-aligned AI Hive image-generation wrapper, with disclosed local API-key setup, user-selected media upload, task polling, and result download behavior.

Install only if you are comfortable giving this skill an AI Hive API key, uploading the reference images you choose to AI Hive/object storage, and saving generated outputs locally. Review the configured output directory and avoid passing sensitive files as references.

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
89% confidence
Finding
The skill documentation invokes a Python script that performs network, shell, environment, and file operations, yet the skill declares no permissions. This creates a transparency and consent problem: users may run the skill without understanding it can store data locally, access environment variables, or make outbound requests, increasing the risk of unintended credential exposure or filesystem changes.

Tp4

High
Category
MCP Tool Poisoning
Confidence
95% confidence
Finding
The documented purpose is narrowly framed as character-consistency image generation, but the referenced implementation appears to include broader capabilities such as API key collection/storage, browser launching, media upload, task polling, and even generic chat/video/model/user-info functions. This mismatch is dangerous because it defeats informed consent and expands the attack surface beyond what a user would reasonably expect from the description.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.