ClawHub
Back to skill

Security audit

Nano Banana 2 图片编辑

Security checks across malware telemetry and agentic risk

Overview

This image-editing skill uploads user-selected images to AI Hive and stores an API key locally, with no evidence of automatic exfiltration, destructive behavior, or hidden execution.

Install only if you are comfortable sending selected source images and prompts to AI Hive and storing an AI Hive API key locally. Use the skill with explicit image paths, avoid the standalone upload command for non-image or sensitive files, and review the local config file if sharing the machine.

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
94% confidence
Finding
The skill invokes shell commands, reads local image files, writes outputs, accesses environment variables via $SKILL_PATH, and performs networked API operations, yet no permissions are declared. This creates a trust-boundary problem: users and orchestrators cannot accurately assess or constrain what the skill can access, increasing the risk of unintended file exposure or external data transfer.

Tp4

High
Category
MCP Tool Poisoning
Confidence
96% confidence
Finding
The skill is presented as a narrowly scoped image-editing tool, but the underlying behavior reportedly includes generic AI Hive/OpenAPI access, chat/multimodal calls, video generation, model enumeration, and account or wallet queries. This mismatch is dangerous because it can cause users or calling systems to grant trust and route tasks under false assumptions, enabling broader data access or unintended actions than advertised.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.