Seedance 视频生成与编辑

• Outcome: Review
• Latest audit: Aug 16, 2026 · 2:59 PM UTC
• Version: 1.0.0

Overview

This appears to be a real AI Hive Seedance video skill, but its activation text is much broader than its video purpose while the workflow uploads media and uses a paid API key.

Review this before installing. Use it only when you intentionally want AI Hive Seedance video generation or editing, and confirm any media upload or paid task submission. Avoid letting it handle generic AI-tool comparisons, e-commerce research, or platform policy questions unless you are explicitly asking it to generate video. Consider using a dedicated AI Hive API key with limited account exposure if available.

SkillSpector

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• MCP Least Privilege: Underdeclared Capability, Wildcard Permission, Missing Permission Declaration
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection

Findings (9)

Lp3

Medium

• Category: MCP Least Privilege
• Confidence: 88% confidence
• Finding: The skill documentation describes capabilities that read environment/config values, write local files, invoke shell commands, and make network requests, but it does not declare permissions accordingly. This weakens user and platform visibility into what the skill can access and do, increasing the chance of unexpected data exposure or local side effects when users provide prompts, media, or API keys.

Tp4

High

• Category: MCP Tool Poisoning
• Confidence: 92% confidence
• Finding: The stated purpose is narrowly framed as Seedance video generation/editing, but the finding indicates the underlying implementation can also perform text chat, image generation, model enumeration, and account/balance retrieval. That mismatch can cause users to authorize or invoke a skill under false assumptions, while the skill accesses broader remote capabilities and potentially more account data than expected.

Description-Behavior Mismatch

Medium

• Confidence: 95% confidence
• Finding: The file implements a broad generic AI Hive client for chat, image, model listing, uploads, and task polling, while the skill is presented as a Seedance video generation/editing skill. This scope mismatch increases attack surface and enables capabilities the user did not intend to grant, which is risky in agent-skill environments where least privilege and tight scoping matter.

Context-Inappropriate Capability

Low

• Confidence: 88% confidence
• Finding: Exposing user-info allows the skill to retrieve account/profile and wallet balance data unrelated to video generation. In an agent context, this expands access to sensitive metadata that could be queried without a clear business need, violating minimization expectations.

Context-Inappropriate Capability

Medium

• Confidence: 94% confidence
• Finding: The skill exposes a generic text chat endpoint that is unrelated to the declared Seedance video-editing purpose. This creates unintended capability expansion, allowing arbitrary model interaction and possible policy bypass or data exfiltration through a channel users would not expect from a video tool.

Context-Inappropriate Capability

Medium

• Confidence: 94% confidence
• Finding: The script exposes generic image generation functionality even though the skill is marketed as Seedance video generation/editing. Unrelated generation features broaden the operational scope and can be abused to perform actions outside the approved purpose, undermining least-privilege design.

Vague Triggers

High

• Confidence: 95% confidence
• Finding: The activation text is extremely broad and includes many generic video, ecommerce, advertising, and competitor-related search terms, making the skill likely to trigger outside its narrow intended use. Overbroad triggering is dangerous because it can route unrelated user requests, prompts, or media into a networked workflow that uploads assets and creates tasks on a third-party service without sufficiently specific user intent.

Vague Triggers

High

• Confidence: 94% confidence
• Finding: The usage guidance repeatedly claims suitability for a very wide set of competitor, platform, marketing, and migration scenarios without defining boundaries. In context, this increases the chance of accidental invocation and silent transmission of user prompts or media to external services in cases where the user may have only been comparing tools or discussing ecommerce workflows.

Missing User Warnings

Medium

• Confidence: 90% confidence
• Finding: The skill says it can automatically upload source media, track tasks, and download outputs, but it does not prominently warn users that their prompts and media are sent to a remote service and that resulting files are saved locally. This is a meaningful transparency and consent issue, especially for potentially sensitive commercial assets, unreleased ads, or personal media.

VirusTotal

VirusTotal findings are pending for this skill version.

Static analysis

No suspicious patterns detected.