Security audit

Amazon 亚马逊电商视频生成与编辑

Security checks across malware telemetry and agentic risk

Overview

This skill is a coherent AI Hive video-generation helper, with user-directed media upload and API-key storage that are expected for its purpose.

Before installing, confirm you are comfortable sending selected product media and prompts to AI Hive, using an AI Hive API key that may incur provider charges, and storing that key locally in ~/.ai-hive/config.json. Use --no-download or --output-dir when you want tighter control over saved outputs.

SkillSpector

By NVIDIA

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 (6)

Lp3

Medium

Category: MCP Least Privilege
Confidence: 91% confidence
Finding: The skill advertises executable CLI workflows that invoke shell commands, read local media files, write outputs, access environment-derived configuration, and make network requests, yet it declares no permissions. This creates a transparency and least-privilege problem: a user or platform may trust the skill as low-risk while it can exfiltrate local files or use networked APIs once executed.

Tp4

High

Category: MCP Tool Poisoning
Confidence: 88% confidence
Finding: The documented purpose is narrowly framed as Amazon video generation/editing, but the analyzed behavior reportedly includes broader chat, image generation, account/wallet retrieval, model catalog access, pricing/routing selection, and generic media upload utilities. This mismatch is dangerous because it expands the attack surface and can enable unexpected data disclosure, unintended charges, or account reconnaissance under the guise of a narrowly scoped media-editing skill.

Description-Behavior Mismatch

Medium

Confidence: 95% confidence
Finding: The file is presented as an Amazon ecommerce video generation/editing skill, but the module docstring and CLI expose broad generic AI Hive capabilities including text chat, image generation, model enumeration, media upload, and account-related operations. Scope expansion like this increases attack surface and enables callers to use the skill for unrelated actions not disclosed by the manifest, undermining least privilege and user expectations.

Description-Behavior Mismatch

Medium

Confidence: 97% confidence
Finding: The skill includes a user-info endpoint that retrieves user information and wallet balance, which is unrelated to Amazon video production. Even if intended for diagnostics, exposing account inspection in a content-generation skill unnecessarily broadens accessible sensitive data and could leak billing or account details to callers or logs.

Description-Behavior Mismatch

Medium

Confidence: 96% confidence
Finding: A standalone text chat feature is outside the declared purpose of an Amazon video generation/editing skill. This creates an undocumented general-purpose LLM capability that can be abused for unrelated content generation, prompt injection experiments, or policy bypass attempts through a skill that users may trust as narrowly scoped.

Description-Behavior Mismatch

Medium

Confidence: 95% confidence
Finding: The file exposes standalone image generation in addition to video generation/editing, exceeding the manifest's stated scope. Undisclosed multimodal capability expansion increases misuse potential and can surprise downstream systems that assume the skill only creates or edits videos for Amazon commerce workflows.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.