Claude Skill MartBrowse skillsQuick linesLearn by videoTerminal guideWhat is a Skill?
 
← Back to list

Seedance Prompt Debugger

Diagnoses 12 common Seedance 2.0 video generation failures and rewrites the prompt to fix them.

Image & VideoIntermediate★ 415⑂ 95AI score 8/10Last updated: Oct 7, 2026

What it does

Give it a broken Seedance 2.0 prompt plus your symptom description (or a frame from the bad video), and it classifies the issue into one of 12 known failure modes, then returns root cause → fixed prompt → a before/after diff table explaining each change.

  • Covers character ID drift, twin/clone duplication, unwanted subtitles, watermarks/logos, style drift (anime turning photoreal), jump cuts when extending clips, quality degradation, wrong VFX, too many reference characters, end-of-clip audio pops, Chinese mispronunciation, and voice-timbre mismatch
  • If you only say "it looks wrong", it insists on three questions first: exact symptom, original prompt, and which reference assets you used
  • For issues prompts cannot fully solve (extension quality loss, timbre), it states the model's known limits and suggests post-production fixes in CapCut/剪映

Who it's for

  • Creators actively shipping short-form or ad videos with Seedance 2.0
  • AI video producers who need character consistency from reference images/audio
  • Teams burning time on trial-and-error prompt edits without finding the cause

Examples

  1. "Two characters arguing, but four identical people appear" → identifies the three-view reference sheet as the cause, adds explicit character↔reference bindings plus a no-clone constraint block
  2. "I asked for 2D anime and got a live-action person" → explains the reference image is the dominant style anchor and proposes a two-step flow: convert the reference to anime first, then generate with cel-shading constraints
  3. "A random shop sign caption appeared bottom-right and a watermark top-left" → adds standard no-subtitle/no-watermark/no-signage-text constraints and recommends landscape over vertical framing

· · · Install guide · · ·

Try it now, no install

Paste this into Claude to use the skill without installing anything.

Read the instructions in this file and follow them to help me:
https://raw.githubusercontent.com/cclank/lanshu-awesome-ai-video-kit/HEAD/skills/seedance-debugger/SKILL.md

What I want: (describe your task here)

If Claude can't open the link, open it yourself and paste the contents instead.

↓ If it works for you, download the ZIP below and install it. Then it runs on its own — no pasting each time.

Install in the Claude app (no terminal)
  1. Download the ZIP with the button below.
  2. In Claude, open Settings → Capabilities and turn on 'Code execution and file creation'. (one time)
  3. Go to Customize → Skills → + → 'Upload a skill' and upload the ZIP.
↓ Download ZIP
Install in Claude Code

Let Claude do it — paste this into Claude Code

Install the skill I found on Claude Skill Mart.
Copy the skills/seedance-debugger folder from the GitHub repo cclank/lanshu-awesome-ai-video-kit into my ~/.claude/skills/seedance-debugger/.
When it's done, tell me in one line what this skill can do.

Install with a command instead

git clone https://github.com/cclank/lanshu-awesome-ai-video-kit.git && mkdir -p ~/.claude/skills && cp -r lanshu-awesome-ai-video-kit/skills/* ~/.claude/skills/

⚠ This is a third-party skill. Check the source repository before installing.

  1. Open a terminal.
  2. Clone the repo: git clone https://github.com/cclank/lanshu-awesome-ai-video-kit.git
  3. Create the skills folder: mkdir -p ~/.claude/skills
  4. Copy the skills in: cp -r lanshu-awesome-ai-video-kit/skills/* ~/.claude/skills/
  5. This skill links to docs under methodology/ and sibling skills (e.g. seedance-prompter), so it is safest to keep the whole repo together, e.g. cp -r lanshu-awesome-ai-video-kit ~/.claude/skills/lanshu-video-kit
  6. Restart Claude Code and try: "My Seedance prompt produced a bad video, help me debug it."
View source on GitHub ↗License: MIT