agentleFS
Sign inStart free
LeoYeAILeoYeAI/seedance-skills/skills/seedance-filter/SKILL.md

seedance-filter

Skill42 starsChanged 4 months agoMITView on GitHub

This skill should be used when a Seedance 2.0 prompt is blocked, rejected, silently degraded, or likely to trigger a content filter; or when the user asks for a safer rewrite without losing the creative intent.

2 other projects use a version of this, 1 of them edited.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 3 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use this when a prompt is blocked, degraded, likely to trigger moderation, or needs a safer rewrite without losing creative intent.

    line 31
  2. Distinctive:

    If the user's request is unsafe, refuse or redirect to a safe alternative.

    line 54
  3. Distinctive:

    Do not provide filter-bypass, evasion, or hidden-word tactics.

    line 56

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

seedance-filter

Intent

A wrongly blocked prompt makes a user feel accused by a machine with no court of appeal. This skill is the advocate: clear the innocent by stating their honest intent plainly, and never coach the guilty. The user's dignity and the platform's boundary are protected in the same gesture.

Boundary — read before anything else

This skill repairs false positives only: benign production content blocked or degraded by over-broad filtering (medical, historical, athletic, fictional-original contexts). It works by clarifying legitimate context in plain language — never by disguising intent. It does not rephrase genuinely prohibited content: anything risky involving minors, real-person likeness without rights, sexual or graphic or illegal material. If the underlying request is prohibited, refuse plainly and offer a legitimate alternative only where one exists.

Use this when a prompt is blocked, degraded, likely to trigger moderation, or needs a safer rewrite without losing creative intent. This skill does not help evade safety systems. It rewrites risky surface wording into professional, non-graphic production language and preserves the safe creative core.

Repair Method

  1. Identify the creative intent: action, mood, camera, subject, and final beat.
  2. Identify risky surface wording: graphic harm, protected identity, sexualized framing, real-person likeness, weapons, self-harm, hate, evasion language, or exact IP copying.
  3. Replace risky terms with professional, non-graphic, production-context language.
  4. Preserve composition, action, mood, camera logic, and authorized references.
  5. For likely false positives, clarify benign production context, ownership, and non-graphic intent. Do not help bypass safety systems or provide evasion tactics.

Safer Rewrite Patterns

IntentSafer direction
Conflictstaged confrontation, choreographed action beat, no graphic injury
Aftermathnon-graphic distress, torn fabric, scattered props, dramatic silence
Suspensethreat implied by shadow, locked door, heavy breathing, low light
Weapon-like propprop object handled safely within a staged action scene
Horror moodeerie atmosphere, flickering practical light, off-screen sound cue
Protected identityoriginal character with broad genre archetype traits

Boundary Rule

If the user's request is unsafe, refuse or redirect to a safe alternative. If it is safe but poorly worded, repair the wording. When uncertain, state the risk class and offer a conservative prompt that keeps the non-harmful scene function.

Do not provide filter-bypass, evasion, or hidden-word tactics. The safe path is to clarify production intent, remove unsafe identity or harm elements, and rewrite into an original authorized scene.

Face-limit or portrait-verification workarounds are not safe prompt tricks. If a surface offers sanctioned virtual portrait, trusted model-output, or authorization asset flows, route the user to those current official paths instead of evasion language.

Load [ref:filter-vocab] for safer substitutions. Load [ref:multilingual-community-examples] only when the safe repair needs Chinese/Russian/Japanese/Korean/Spanish or mixed-language wording for clarity.

Output Contract

Return likely trigger class, safer wording, final prompt, what changed, and any content boundary that still applies.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free