seedance 2.0 guide

• Outcome: Review
• Latest audit: May 28, 2026, 11:53 AM
• Version:
  1.0.5

Overview

This is mostly a creative video prompt guide, but it includes advice that appears aimed at avoiding face-related safety checks.

Review this skill before installing. It appears safe from a local-system security perspective, but its prompt guidance around faces and identity could be used to work around downstream safety systems. Only use media you have rights to use, avoid prompts involving unauthorized likenesses, and do not rely on stylization, framing, or wording changes to bypass platform rules.

SkillSpector

By NVIDIA

Vulnerability Patterns

• 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
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (1)

Ssd 2

Medium

• Confidence: 96% confidence
• Finding: This section explicitly advises users to rephrase prompts and alter framing to avoid face/identity-focused safety checks, including obscuring faces, using stylized substitutes, and replacing realistic biometric descriptors with indirect artistic wording. That is not neutral creative guidance; it meaningfully facilitates policy evasion for identity/face-related safeguards in downstream image/video generation systems.

VirusTotal

64/64 vendors flagged this skill as clean.