Nano Banana Pro Image Gen(基于API易代理站)

• Outcome: Pass
• Latest audit: May 28, 2026, 11:00 AM
• Version:
  0.1.2

Overview

This is a legitimate image-generation skill that sends prompts and selected images to APIYI, so users should be mindful of privacy but no hidden or malicious behavior was found.

Install only if you are comfortable sending image prompts and any chosen reference/edit images to APIYI and downstream model services. Prefer the APIYI_API_KEY environment variable over passing keys on the command line, and do not include secrets, private documents, or sensitive images unless that external processing is acceptable.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (4)

Missing User Warnings

Medium

• Confidence: 84% confidence
• Finding: The skill sends prompts and potentially user-supplied images to a third-party domestic proxy API and writes files locally, but does not clearly warn about privacy, retention, or transmission risks. Users may unknowingly submit sensitive text or images to an external processor, especially because the skill encourages use of raw user input.

Missing User Warnings

Medium

• Confidence: 93% confidence
• Finding: The script transmits user prompts and optionally up to 14 local images to a third-party API endpoint, but it does not present a clear privacy warning or require explicit consent at runtime. In a skill context, users may assume processing is local or trusted, causing accidental disclosure of sensitive images, embedded metadata, or confidential prompt content to an external service.

Missing User Warnings

Medium

• Confidence: 86% confidence
• Finding: The script transmits user prompts and base64-encoded input images to a remote third-party API, but it does not present an explicit user-facing warning or consent step at runtime. In this skill context, users may provide sensitive images or confidential prompt content, so silent transfer to an external service creates a real privacy and data-handling risk.

Ssd 3

Medium

• Confidence: 97% confidence
• Finding: The skill explicitly instructs the agent to forward the user's full raw input directly to the external image-generation service. This can leak unnecessary sensitive data, including personal information, credentials, unrelated context, or system artifacts accidentally present in the user's message, to a third-party API.

VirusTotal

66/66 vendors flagged this skill as clean.