Seedance CopyrightSAFE

Comprehensive production pipeline for quad-modal AI filmmaking with Seedance 2.0

The instruction file, verbatim from the audited commit — this is the text the model reads, and the surface the audit's instruction layer examines. Quoted here so you can judge it without cloning anything.

---
name: seedance-copyright
description: "This skill should be used when a Seedance 2.0 prompt mentions named characters, franchises, studios, celebrities, public figures, private people, brand logos, copyrighted scenes, songs, voices, or real-person likeness workflows and needs an IP-safe rewrite."
license: MIT
user-invocable: true
tags:
  - copyright
  - ip
  - likeness
  - safety
  - seedance-20
metadata:
  version: "6.7.0"
  updated: "2026-08-01"
  parent: "seedance-20"
  author: "Iamemily2050 (@iamemily2050)"
  repository: "https://github.com/Emily2040/seedance-2.0"
  openclaw:
    emoji: "🎬"
    homepage: "https://github.com/Emily2040/seedance-2.0"
---

# seedance-copyright

Before producing prompt text, a prompt-ready block, a rewrite, an example, or a compiled clip, load the [Director's Read](../../references/directors-read.md), classify the brief, and complete its canonical narrative or non-narrative record. Translate that record into visible or audible carriers and keep its internal labels out of final generation prose.

Use this before finalizing prompts involving protected IP, named brands, public figures, private people, voices, logos, songs, studios, exact scenes, or lookalike character requests. The goal is not to dilute the idea; the goal is to preserve the creative function with original, authorized, and safer production language.

## Intent

The user pointing at protected work is not trying to steal - they are showing you the clearest example of what they love that exists. The job is to find what the love is made of and give it back as something safely theirs. Protect the rights-holder and the user's heart in the same move.

## Rewrite Principle

Preserve the scene function, genre, mood, camera logic, emotional beat, and production intent. Replace protected identity with an original archetype, original costume logic, original world details, and descriptive style layers.

| Risk | Replace with |
|---|---|
| Named character or franchise | Original archetype, genre function, and non-identical costume language |
| Studio or living-creator style | Medium, texture, palette, composition, line quality, and motion rhythm |
| Celebrity or private person | Original performer description or authorized reference workflow |
| Brand logo | Generic product mark, blank label, or user-owned brand if explicitly authorized |
| Song, voice, or performance | Tempo, energy, instrumentation, mood, or newly composed sound direction |
| Exact scene recreation | Original scene with similar narrative function and different setting/blocking |

## Authorization Gate

If the user clearly owns the brand, asset, or likeness rights, keep the authorized elements but still preserve them with explicit constraints. If authorization is unclear, ask a short confirmation or provide a safe original rewrite. Do not assume rights from an uploaded image, song, or video.

For real human faces, portraits, or voices, separate three questions: does the active surface support the input, does the user have authorization, and does the prompt avoid imitation of a public figure or private person without consent. Some surfaces use verified virtual portrait assets or authorization flows; do not collapse those into a universal allow or deny rule.

## Safe Replacement Example

Instead of a named superhero swinging through a recognizable franchise city, write: `original masked rooftop courier in a red weatherproof jacket leaps between rain-slick buildings, low handheld tracking camera, blue police lights far below, no logos or franchise symbols`.

## Output Contract

Return risk category, what was changed, safe replacement prompt, authorization requirement, and any residual constraints.

SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.

What the source does

Filesystem

none-observed

Network

none-observed

Shell

none-observed

Dependencies

pinned

Secrets in source

none-found

Findings (0)

No findings outside the package's declared scope.

Gates applied: no_behavioural_pass.