小红书 电商视频生成与编辑

• Outcome: Pass
• Latest audit: Aug 17, 2026 · 6:08 AM UTC
• Version: 1.0.0

Overview

This skill is a coherent AI Hive video-generation helper that uses user-supplied media and an API key, with some extra dormant generic client code that is not exposed by the normal CLI entrypoint.

Before installing, understand that this skill will send prompts and any media files you select to AI Hive, may incur provider costs, and stores an AI Hive API key in your home directory if you run init. Avoid giving it unrelated private files, and review AI Hive account/API-key permissions if you share the machine.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• MCP Least Privilege: Underdeclared Capability, Wildcard Permission, Missing Permission Declaration
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration

Findings (6)

Lp3

Medium

• Category: MCP Least Privilege
• Confidence: 84% confidence
• Finding: The skill advertises executable commands that install packages and run a Python script with shell, network, file read/write, and environment access, yet it declares no permissions. This undermines least-privilege review because users and automated policy systems cannot accurately assess what the skill can do before execution.

Tp4

High

• Category: MCP Tool Poisoning
• Confidence: 90% confidence
• Finding: A skill presented as a narrow Xiaohongshu ecommerce video tool appears to expose broader AI Hive/OpenAPI functions such as general chat, image generation, model enumeration, and account or wallet queries. This scope expansion is dangerous because users may grant trust or invoke the skill for a limited purpose while it can access unrelated functionality and potentially sensitive account information.

Description-Behavior Mismatch

Medium

• Confidence: 93% confidence
• Finding: The file is packaged as a focused 小红书电商视频生成/编辑 skill, but it contains a broader generic AI Hive client exposing chat, image generation, model enumeration, uploads, and account-related operations. This expands the capability surface beyond what a caller would reasonably expect from the skill, increasing the chance of misuse, policy bypass, or unintended data access.

Context-Inappropriate Capability

Medium

• Confidence: 88% confidence
• Finding: The skill includes a user-info endpoint that retrieves account information and wallet balance, which is unrelated to the declared ecommerce video-generation purpose. In a skill ecosystem, unnecessary access to billing/account metadata violates least privilege and can expose sensitive account details to workflows that only need media generation.

Context-Inappropriate Capability

Medium

• Confidence: 90% confidence
• Finding: The standalone text-chat capability is outside the declared scope of a video-generation/editing skill. That broader interface can be repurposed for unrelated model access, prompt handling, or data transmission, creating an unnecessary and potentially confusing attack surface for consumers expecting only video functionality.

Context-Inappropriate Capability

Medium

• Confidence: 90% confidence
• Finding: The file exposes standalone image-generation functionality even though the skill is presented as a focused video-generation/editing tool. This mismatch broadens accessible model operations and can enable out-of-scope content generation or cost-incurring actions not anticipated by users or platform reviewers.

VirusTotal

65/65 vendors flagged this skill as clean.

Static analysis

No suspicious patterns detected.