Security audit

Seedance 参考视频原创复刻｜AI-HIVE

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed AI-HIVE video-generation workflow that uses user-provided media and an API key for its stated purpose, with some review notes around credential storage and auto-invocation.

Before installing, confirm you are comfortable providing an AI-HIVE API key, uploading only media you are authorized to use, and reviewing parameters before any generation that may incur cost. Use environment variables or the local config file deliberately, and avoid storing real API keys in logs or repositories.

SkillSpector

Vulnerability Patterns

Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration

Findings (4)

Lp3

Medium

Category: MCP Least Privilege
Confidence: 87% confidence
Finding: The skill embeds executable examples and operational guidance that use environment variables, local files, network access, and shell tools, but it does not declare any corresponding permissions. This creates a transparency and consent gap: users or the hosting platform may invoke a skill with broader capabilities than expected, increasing the risk of unintended file access, outbound requests, or local command execution.

Tp4

High

Category: MCP Tool Poisoning
Confidence: 92% confidence
Finding: The documented purpose is a constrained, originality-focused reference-video remake workflow, but the static analysis indicates materially broader behavior: generic text/image generation, model enumeration, wallet/user-info access, local credential storage, and general-purpose ffmpeg editing. This mismatch is dangerous because it can mislead users and policy systems about what the skill can actually do, enabling overbroad data access, unexpected billing exposure, and use as a more general execution or content-generation tool than advertised.

Vague Triggers

Medium

Confidence: 88% confidence
Finding: The manifest allows implicit invocation, which means the skill can be triggered without an explicit user request naming the skill or clearly confirming its use. In a skill that can generate production-ready workflows and runnable AI-HIVE commands for reference-video remake tasks, broad auto-activation increases the chance of unintended execution in sensitive copyright, marketing, or policy-bound contexts.

Natural-Language Policy Violations

Medium

Confidence: 79% confidence
Finding: The default prompt hard-codes Chinese output regardless of user preference, which can reduce transparency and informed review if the user expected another language. While not a direct code-execution issue, forced language selection can obscure risky instructions, consent boundaries, or copyright-related caveats in a workflow that handles reference-video remake requests.

Static analysis

No suspicious patterns detected.