Seedance 2.0 — AI Video by ByteDance

• Outcome: Review
• Latest audit: May 29, 2026 at 9:25 PM UTC
• Version:
  1.1.2

Overview

This appears to be a real Atlas Cloud video-generation helper, but it needs review because its script can use broader paid model and media-upload capabilities than the main skill description clearly discloses.

Install only if you intend to use Atlas Cloud for paid video generation. Use a dedicated API key, avoid confidential prompts or private media, confirm which model and files will be sent before running commands, and treat the upload command as sending local files to Atlas Cloud.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• 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
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access

Findings (2)

Tp4

High

• Category: MCP Tool Poisoning
• Confidence: 88% confidence
• Finding: The finding indicates the implemented behavior is broader than the declared purpose: arbitrary Atlas Cloud video models, additional generation modes, model listing, and file upload capabilities are reportedly supported despite the skill being presented as a narrowly scoped Seedance 1.5 skill. That mismatch is dangerous because users and host systems may grant trust, routing, or data access based on the narrower description, while the actual skill can send more data types to external services or invoke unintended model behaviors.

Vague Triggers

Medium

• Confidence: 84% confidence
• Finding: The trigger text is extremely broad, covering essentially any AI video, marketing video, reel, talking head, or video-content request. Overbroad invocation can cause the skill to activate in contexts where users did not intend external transmission of prompts, image URLs, or media to Atlas Cloud, increasing the chance of privacy leakage, unnecessary cost, and misuse of the API key.

VirusTotal

66/66 vendors flagged this skill as clean.