TikTok Shop 爆款带货视频生成

• Outcome: Pass
• Latest audit: Aug 17, 2026 · 2:52 AM UTC
• Version: 1.0.0

Overview

This skill is a disclosed AI Hive video-generation helper for TikTok Shop creative workflows, with expected API-key setup, media upload, task polling, and result downloads.

Install only if you are comfortable giving the helper an AI Hive API key and uploading the product photos, videos, or audio you choose to provide. Generated outputs are downloaded locally by default to ~/Downloads/AiHive unless you set a different output directory.

SkillSpector

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: 89% confidence
• Finding: The skill exposes effective capabilities for environment access, filesystem read/write, network calls, and shell execution without declaring permissions. That creates a transparency and containment gap: users and policy systems may invoke a seemingly simple video-generation skill without realizing it can access local data or execute arbitrary commands through helper scripts.

Tp4

High

• Category: MCP Tool Poisoning
• Confidence: 95% confidence
• Finding: The declared purpose is narrow, but the underlying behavior appears to include broad AI Hive chat, image generation, account/balance lookup, model enumeration, upload utilities, and initialization flows unrelated to the stated TikTok sales-video use case. This mismatch is dangerous because it expands the attack surface and can enable unexpected data exfiltration, credential use, or repurposing of the skill beyond what a user reasonably consented to.

Description-Behavior Mismatch

Medium

• Confidence: 88% confidence
• Finding: The file presents itself as a TikTok Shop sales-video skill, but the module is actually a general-purpose AI Hive client that supports chat, image generation, model enumeration, uploads, and task operations. This scope mismatch increases the attack surface and enables users or downstream agents to invoke capabilities far beyond the declared purpose, undermining least-privilege and trust boundaries for the skill.

Context-Inappropriate Capability

Medium

• Confidence: 91% confidence
• Finding: The skill exposes a text chat command even though the declared purpose is TikTok Shop video generation. Unrelated LLM access can be abused for prompt processing, data exfiltration through model calls, or policy bypass in environments that only intended to permit constrained media generation.

Context-Inappropriate Capability

Medium

• Confidence: 90% confidence
• Finding: Standalone image generation is outside the stated sales-video scope and broadens what this skill can do without clear user expectation. That creates unnecessary privilege and may let an agent pivot from constrained video creation into arbitrary image generation workflows.

Context-Inappropriate Capability

Low

• Confidence: 78% confidence
• Finding: User-info and model-listing commands reveal account and platform metadata that are not necessary for generating TikTok Shop videos. While not directly code-execution, they expand reconnaissance capability and can disclose wallet, account, or available-model details to an actor who only needed video generation.

Static analysis

No suspicious patterns detected.