Agent skill

Scenario Seedance Storyboard

by scenario-labs in scenario-labs/skills

A skill your agent uses when a Seedance video must hold continuous movement across shots: dance choreography, fights, sports, or any performance where limbs teleport, feet slide, or motion resets at…

MITAuto-check passedMedia & Creative

Install Scenario Seedance Storyboard

skills CLI
$ npx skills add scenario-labs/skills --skill scenario-seedance-storyboard -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install scenario-labs/skills scenario-seedance-storyboard --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/scenario-labs/skills.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/scenario-seedance-storyboard .claude/skills/scenario-seedance-storyboard && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
scenario-seedance-storyboard
GitHub stars
946
Token cost
~2.7k tokens
SKILL.md length
1,424 words
Files
5 (incl. references)
Skills in repo
146
Repo updated
First seen
Licence
MIT

At a glance

A skill your agent uses when a Seedance video must hold continuous movement across shots: dance choreography, fights, sports, or any performance where limbs teleport, feet slide, or motion resets at…

  • Works in 6 steps: Write the script: 12 shots over 24… → Generate one plate per character, then… → search with target="models",… → …
  • A Seedance video must hold continuous movement across shots: dance choreography
  • SKILL.md covers Overview, Quick reference, Two lanes and Worked example: a 12-shot…, plus 1 more section
  • Calls npx

What it does

Scenario Seedance Storyboard is an agent skill from scenario-labs/skills. Use when a Seedance video must hold continuous movement across shots: dance choreography, fights, sports, or any performance where limbs teleport, feet slide, or motion resets at every cut, or when turning a storyboard and character sheet into a multi-shot sequence that plays as one take. Keywords: storyboard to video, choreography, movement control, pose chain, timecoded shot script, first and last frame chaining, character sheet, multi-shot continuity, Seedance 2.5 and 2.0.

Its SKILL.md is about 2.7k tokens, which your agent loads only when the skill is triggered. The skill folder holds 5 other files, including reference files (for example `README.md`, `references/board-craft.md` and `references/chained-lane.md`).

It sits in Media & Creative, covering Comics and storyboards, AI video generation and Slides and decks. It works with Seedance. The repository describes itself as: Get production-ready images, video, audio, and 3D from any AI agent: skills that pick the right model, price before spending, and keep characters and brands consistent through… The licence is MIT.

When your agent uses it

  • A Seedance video must hold continuous movement across shots: dance choreography
  • Any performance where limbs teleport
  • Motion resets at every cut
  • Turning a storyboard and character sheet into a multi-shot sequence that plays as one take

Example prompts

  • “/scenario-seedance-storyboard”

Requirements

  • Node.js

Workflow steps

6 steps, taken from the first numbered list in SKILL.md.

  1. Write the script: 12 shots over 24 seconds, edge poses named, every exit matching the next entry. Show it to the user first; it is the…
  2. Generate one plate per character, then the 12 panels, then two or three photoreal key frames from the same plates. Upload anything local…
  3. search with target="models", query="seedance", public=true. Hits rank by relevance, not generation, so scan them all for the newest…
  4. model_run with dry_run=true and parameters={"prompt": "", "referenceImages": ["asset_dancer", "asset_board", "asset_key1", "asset_key2"]…
  5. Re-run with wait=false, then jobs_wait, re-called with pending_job_ids on timeout. A timeout is not a failure and never justifies a second…
  6. Step through the cuts: both sides must agree in pose and match the panel. Then sweep the whole clip at two frames a second, because a…

What it can do on your machine

Read from SKILL.md and the folder at commit f6f8ab7. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Shell commands in SKILL.md call:

    • npx

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md. Its commands use npx, which can reach the network depending on how they are called.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Scenario Seedance Storyboard loads about 2.7k tokens when it runs, and up to ~8.3k if it reads all its reference files. Until then it costs about 127 tokens; SKILL.md has 1,424 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~127
When it runs · the whole SKILL.md, loaded when a task matches
~2.7k
With references · SKILL.md plus every file in references/, read only if the agent opens them
~8.3k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from scenario-labs/skills at commit f6f8ab7, republished under its MIT licence (© scenario-labs). 1,424 words, ~2,709 tokens.

Download SKILL.mdSave it as .claude/skills/scenario-seedance-storyboard/SKILL.md (or your agent's skills folder). This skill also uses 4 other files; get the full folder from GitHub.
name
scenario-seedance-storyboard
description
Use when a Seedance video must hold continuous movement across shots: dance choreography, fights, sports, or any performance where limbs teleport, feet slide, or motion resets at every cut, or when turning a storyboard and character sheet into a multi-shot sequence that plays as one take. Keywords: storyboard to video, choreography, movement control, pose chain, timecoded shot script, first and last frame chaining, character sheet, multi-shot continuity, Seedance 2.5 and 2.0.
license
MIT

Scenario Seedance Storyboard

Overview

Prompted shot by shot, generated movement falls apart: limbs teleport, feet slide, and every cut resets the performance. So storyboard instead of prompting harder. What this delivers is continuous motion through the handoffs, not editorially distinct cuts: for cuts a viewer can feel, cut the delivered clip yourself with scenario-video-editing. Write a timecoded shot script, draw a character sheet and a numbered panel per shot, and hold one rule: every shot ends in the pose the next one starts from. Pinned boundaries leave nothing to reset, so the cuts play as one performance.

Connection and the core loop: the scenario skill; the Seedance parameter contract, conditioning traps, and sound lanes: the scenario-seedance skill; identity across stills and a reusable character sheet: the scenario-consistency and scenario-identity-library skills; a board or comic page that is itself the deliverable: the scenario-storyboards skill; splitting and rejoining shots: the scenario-video-editing and scenario-video-assembly skills. If a sibling skill named here is missing from your available skills, ask the user to install it (npx skills add scenario-labs/skills --skill <name>); unattended, proceed from tool schemas and flag the gap.

Quick reference

StepWhatHow
1. Scripttimecoded list, cast and setting agreed firstper shot: number, size, angle, lens or move, action verbs, entry and exit poses
2. Platesone image per performer who must stay recognizable, several angles; extras generic and unnamedimage model via recommend; one clean still of a real performer also works
3. Briefsone line per panel, coverage not picturesmaster first, hold the axis, climb the size ladder, punctuate
4. Panelsone model_run per panelplates as references, light pencil line on bare paper, each at delivery ratio
5. Gatetwo passes per panel, then repair one panelartifacts and differences, per board-craft
6. Assemblethe approved panels into the boardmodel_scenario-compose-image at explicit coordinates
7. Approvescript, plates, board shown to the userreshoot by panel number; unattended, record and continue
8. Framestwo or three photoreal key framessame plates, delivery ratio; they set the delivered look
9. Videoone scripted run, or chained runsSeedance model_run, one of the two lanes below
10. Checkthe delivery is clean and carries its tracksample every half second, not only the cuts, and match each cut to its panel

Label each panel as a shot list does: number, shot size, angle, lens or camera move, then three action verbs. Arrange them as coverage, not twelve pictures: open on a master that sets the geography, hold one side of the axis so screen direction never flips, climb the size ladder, and punctuate with an insert or a reaction. Adjacent shots share an edge by construction: shot 3's exit is shot 4's entry.

Draw the board light: thin pale line on bare paper, the setting in two or three quick lines, no rendered environment. A board drawn as a competing picture is treated as one, and its markings render into the delivery as objects, which no prompt clause can clean. Drawn light, the same markings transfer nothing, so numbers, captions and arrows can all stay on the board a human reviews. What carries that is prompt wording, not drawing: declare the marks as the director's planning guides rather than prohibiting them, and pass the key frames as concept art so appearance has a photographic source. Both are in video-prompt.md.

Panels are generated one at a time, so one bad panel costs one panel, and gated before any video run on two passes: an AI-quality score for artifacts, and a spot-the-difference against the plate. A board-level verdict is not the bar: panels average defects away.

File as you go: one collection per sequence, created before the first generation, holding plates, panels, repairs, board and video (board-craft).

Stop for approval before the first video run, with the key frames shown alongside the script, plates and board, and invite a reshoot by panel number; unattended, record the assumption and continue.

Two lanes

One scripted run. When the sequence fits one job's duration cap (30 seconds on 2.5 at authoring time), pass the plates, board and key frames as referenceImages and build the prompt to the seven-section shape in video-prompt.md. One generation carries the internal cuts: the plates hold identity, the pose chain holds continuity. Name aspectRatio; it defaults to adaptive here, so a portrait board hands back a portrait video. Reference mode anchors frame one to the base state of the reference world (see scenario-seedance), so a sequence that must open on an exact pose belongs in the chained lane.

Chained per-shot runs. Past the duration cap, when each shot needs its own control, or when the sequence must open on an exact pose: render each boundary pose as a still, then run shot N with image = still N-1 and lastFrameImage = still N, so adjacent shots share a frame both sides agree on. This excludes referenceImages, so the stills are a shot's only identity, and N shots cost N+1 stills and N runs. Each run also has a duration floor (4 whole seconds on 2.5 at authoring time), so N shots cannot total less than N times it, and more cuts than that means fewer shots or more seconds. The disciplines and the repair path are in references/chained-lane.md.

Show full SKILL.md (548 more words)Show less

Worked example: a 12-shot dance in one run

  1. Write the script: 12 shots over 24 seconds, edge poses named, every exit matching the next entry. Show it to the user first; it is the cheapest place to be wrong; unattended, write it down and continue.
  2. Generate one plate per character, then the 12 panels, then two or three photoreal key frames from the same plates. Upload anything local with upload_asset (see the scenario skill). Show the board and the key frames to the user before going further.
  3. search with target="models", query="seedance", public=true. Hits rank by relevance, not generation, so scan them all for the newest non-deprecated member rather than the first: at authoring time model_bytedance-seedance-2-5 (re-discover each session), which takes the most references and holds identity best. Then model_schema_get for caps.
  4. model_run with dry_run=true and parameters={"prompt": "<the timecoded script>", "referenceImages": ["asset_dancer", "asset_board", "asset_key1", "asset_key2"], "duration": 24, "resolution": "1080p", "aspectRatio": "16:9"}; re-estimate after any change.
  5. Re-run with wait=false, then jobs_wait, re-called with pending_job_ids on timeout. A timeout is not a failure and never justifies a second model_run.
  6. Step through the cuts: both sides must agree in pose and match the panel. Then sweep the whole clip at two frames a second, because a leaked board marking can fade in part-way through a shot, so one sample per shot reports a false clean. asset_display returns a link for video, not frames, so pull them locally or with model_scenario-video-to-image-seq (a fixed id: Scenario's single deterministic frame-extraction tool, so discovery would only re-derive it). Expect fewer cuts than scripted: most handoffs render as camera moves, so cut the delivered take at the twelve scripted points.

Common mistakes

  • Prompting choreography as adjectives ("dances energetically"): timecoded verbs and edge poses survive generation; mood words do not.
  • Naming the pose but not the position: in a two-character scene a mirrored boundary still breaks the cut as hard as a wrong limb.
  • Scripting a character who leaves frame and returns, or a change made off camera, inside one run: whatever leaves frame is gone for the rest of that run, and no boundary check can catch what happened out of shot. Land departures, returns and state changes on a chained boundary still (chained-lane).
  • Expecting a scripted run to open on a prompt-described pose: reference mode anchors frame one to the base state; exact openings are the chained lane's job.
  • Delivering silence by default: ask for diegetic sound by naming its sources, and name no instrument, genre, or exclusion such as "no music", which is what triggers an audio refusal. generateAudio: false is the fallback, not the default.
  • A script past the prompt's max_length, or timecodes past the duration cap: read both off model_schema_get before writing the script, since step 1 comes before discovery; an overrun is a 400, never a trim. Then send the run's scripted length as duration: its default is Auto (-1), which the schema defines only against a reference clip, so a timecoded script sent without it runs for an unstated length.
  • Getting the ratios wrong: each panel must carry the delivery ratio, since a reframe moves the poses the chain depends on, and the composed board must land inside a 0.4 to 2.5 aspect ratio or Seedance rejects it as a reference and the run fails.

© scenario-labs, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 4 other files (references) in skills/scenario-seedance-storyboard of scenario-labs/skills.

  • SKILL.md
  • README.md
  • references/board-craft.md
  • references/chained-lane.md
  • references/video-prompt.md

Open the folder on GitHubat commit f6f8ab7

Compare with similar skills

Scenario Seedance Storyboard next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Scenario Seedance Storyboard compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Scenario Seedance Storyboard this skillscenario-labs/skills946—~2.7kAutomated safety check: PassMIT
Scroll Promo Site Builderkangarooking/kangarooking-skills662—~2.6kAutomated safety check: PassMIT
Seedance 2.0 Storyboard Promptselementsix/elementsix-skills311—~1.4kAutomated safety check: PassMIT
Character Design with genmediafal-ai-community/skills251—~1.3kAutomated safety check: PassNone
Painterly 3D2 Cinema Promptsliucongg/liucong-skills248—~1.8kAutomated safety check: PassApache-2.0
Seedance 2.5 Video PlannerAtlasCloudAI/awesome-seedance-2.5-prompts-skills234—~4.3kAutomated safety check: PassCustom licence

Similar skills

  • Scroll Promo Site Builder

    kangarooking/kangarooking-skills

    Create a scroll-controlled cinematic product website with rich motion (动效网站) from product materials, reference pages or videos, and brand assets.

    662 GitHub stars~2.6k tokensUpdated 1 mo ago
    Media & CreativeAuto-check passed
  • Seedance 2.0 Storyboard Prompts

    elementsix/elementsix-skills

    Interviews you about a video idea, then writes a timed shot-by-shot prompt for Seedance 2.0 using its image, video and audio reference syntax.

    311 GitHub stars~1.4k tokensUpdated 7 mo ago
    Media & CreativeAuto-check passed
  • Character Design with genmedia

    fal-ai-community/skills

    Builds consistent characters, reference and expression sheets, outfit variations and character-to-video shots with the genmedia CLI while keeping identity stable.

    251 GitHub stars~1.3k tokensUpdated 12 days ago
    Media & CreativeAuto-check passed
  • Painterly 3D2 Cinema Prompts

    liucongg/liucong-skills

    Expands one sentence into a prompt package for painterly 3D-to-2D short films: story, 15-second segments, characters, scenes, Midjourney storyboards and Seedance video prompts.

    248 GitHub stars~1.8k tokensUpdated 1 mo ago
    Media & CreativeAuto-check passed
  • Seedance 2.5 Video Planner

    AtlasCloudAI/awesome-seedance-2.5-prompts-skills

    Plans and generates controllable Seedance videos by choosing a creative route first, from text prompts and storyboards to keyframe pairs, extensions and edits.

    234 GitHub stars~4.3k tokensUpdated today
    Media & CreativeAuto-check passed
  • Seedance Production Workflow

    LearnPrompt/awesome-seedance

    Plan and carry a Seedance video from brief through references, storyboard, generation handoff, shot-level review, and targeted repair.

    1.9k GitHub stars~1.5k tokensUpdated today
    Media & CreativeAuto-check passed

More from scenario-labs/skills

All 146 skills in this repo
  • Scenario Blender Grease Pencil

    scenario-labs/skills

    A skill your agent uses when drawing or animating with Grease Pencil in Blender 5.x from Python: 2D or 2.5D illustration, frame-by-frame animation, a cutout or part-based 2D character, strokes with…

    946 GitHub stars~4.5k tokensUpdated yesterday
    Auto-check passed
  • Scenario Blender Hair

    scenario-labs/skills

    A skill your agent uses when grooming hair or fur in Blender with hair curves, such as a character hairstyle, animal fur, procedural fur in geometry nodes, or hair cards and mesh hair for games.

    946 GitHub stars~4.7k tokensUpdated yesterday
    Auto-check passed
  • A skill your agent uses when lighting, rendering or compositing in Blender: light a character, product or hero shot, interior at dusk or night, three-point or motivated lighting, sun and sky, HDRI…

    946 GitHub stars~5k tokensUpdated yesterday
    Auto-check passed
  • Scenario Chatgpt Pet Create

    scenario-labs/skills

    A skill your agent uses when creating a ChatGPT pet or Codex pet with Scenario: hatching an animated companion from a text idea, a character, mascot or brand cue, or reference photos and art; making…

    946 GitHub stars~3.6k tokensUpdated yesterday
    Auto-check passed
  • Scenario Godot Animation

    scenario-labs/skills

    A skill your agent uses when animating characters or scenes in Godot 4.7: AnimationPlayer clips and RESET, AnimationTree state machines and blend spaces built in code, Mixamo or glTF import, loop…

    946 GitHub stars~4.7k tokensUpdated yesterday
    Auto-check passed
  • Scenario Godot Audio

    scenario-labs/skills

    A skill your agent uses when adding or fixing sound in Godot 4.7: audio buses and effects, volume sliders, 'too many sounds', combat audio with hundreds of enemies, sounds clipping or distorting, 3D…

    946 GitHub stars~4.7k tokensUpdated yesterday
    Auto-check passed

Works with

Questions about Scenario Seedance Storyboard

What does Scenario Seedance Storyboard do?

A skill your agent uses when a Seedance video must hold continuous movement across shots: dance choreography, fights, sports, or any performance where limbs teleport, feet slide, or motion resets at…. Scenario Seedance Storyboard is an agent skill from scenario-labs/skills. Use when a Seedance video must hold continuous movement across shots: dance choreography, fights, sports, or any performance where limbs teleport, feet slide, or motion resets at every cut, or when turning a storyboard and character sheet into a multi-shot sequence that plays as one take.

When should I use Scenario Seedance Storyboard?

Scenario Seedance Storyboard fits situations like: A Seedance video must hold continuous movement across shots: dance choreography; any performance where limbs teleport; motion resets at every cut; turning a storyboard and character sheet into a multi-shot sequence that plays as one take.

How do I install Scenario Seedance Storyboard in Claude Code?

Run `npx skills add scenario-labs/skills --skill scenario-seedance-storyboard -a claude-code`. Or copy the skill folder (skills/scenario-seedance-storyboard in scenario-labs/skills) into .claude/skills/scenario-seedance-storyboard in your project. Claude Code loads it when a task matches its description.

How do I install Scenario Seedance Storyboard in Codex?

Run `npx skills add scenario-labs/skills --skill scenario-seedance-storyboard -a codex`. Or copy the skill folder (skills/scenario-seedance-storyboard in scenario-labs/skills) into .agents/skills/scenario-seedance-storyboard in your project. Codex loads it when a task matches its description.

Can I use Scenario Seedance Storyboard in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add scenario-labs/skills --skill scenario-seedance-storyboard -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/scenario-seedance-storyboard, .gemini/skills/scenario-seedance-storyboard, .github/skills/scenario-seedance-storyboard and .opencode/skills/scenario-seedance-storyboard in your project.

What does Scenario Seedance Storyboard need to run?

Going by SKILL.md and its folder, Scenario Seedance Storyboard needs the command-line tools its instructions call (npx). Our summary lists: Node.js.

Does Scenario Seedance Storyboard access the network?

SKILL.md contains no URLs. Its commands use npx, which can reach the network depending on how they are called. This is read from the text; nothing was executed.

Is Scenario Seedance Storyboard safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Scenario Seedance Storyboard use?

Scenario Seedance Storyboard is published under the MIT licence (declared in SKILL.md). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Scenario Seedance Storyboard use?

About 2.7k tokens (SKILL.md is roughly 11k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 5.6k tokens, read only when the agent opens those files.

What are the alternatives to Scenario Seedance Storyboard?

Skills that share tags, products or a category with Scenario Seedance Storyboard: Scroll Promo Site Builder (kangarooking/kangarooking-skills, 662 stars), Seedance 2.0 Storyboard Prompts (elementsix/elementsix-skills, 311 stars), Character Design with genmedia (fal-ai-community/skills, 251 stars) and Painterly 3D2 Cinema Prompts (liucongg/liucong-skills, 248 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Scenario Seedance Storyboard?

scenario-labs (a GitHub organization) maintains it in scenario-labs/skills, which has 946 GitHub stars. The repository holds 146 skills in this directory. The repository was last updated on October 10, 2026.

Source: scenario-labs/skills on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.