Seedance 2.5 Skill
A skill that improves Seedance prompts, plans storyboards and generates video through Atlas Cloud, plus a library of ready prompts
- 260 ★ on the repository
- Last commit
Extended access
Why this level
- Generation requests are paid and go to Atlas Cloud with your API key
- References and prompts are sent to the provider's server
- The skill runs the generate.mjs script on your machine
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/seedance-2-5-skillDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- Generation requests are paid and go to Atlas Cloud with your API key
- References and prompts are sent to the provider's server
- The skill runs the generate.mjs script on your machine
This is third-party code. Review the repository files before installing.
What it does
Seedance 2.5 Skill is an installable skill for working with ByteDance's Seedance video models. It takes a creative brief, an existing prompt, references or a storyboard and brings it to a production-ready prompt, prepares the assets, sends the request and checks the result. First the skill picks a route: a short scene from text, a multi-shot sequence from a storyboard, a clip from a set of references, a shot between given first and last frames, an extension, an edit of part of finished video or a transition between two clips. It makes a multi-panel storyboard with Seedream 5.0 Pro, and only when planning or consistency of characters is needed. The image is shown as progress and the skill reviews it itself for panel order and recognizable characters. Video runs through Seedance 2.0 today, and Seedance 2.5 switches on when the selected provider really exposes that model. Task status is polled every 2 seconds on one id, and on a timeout the skill never submits a second paid request. The same repository holds the Universal Video Prompt Skill, which writes one prompt spec and compiles it to the dialect of whichever model you can call, and a library of more than a hundred prompts with sources, authors and previews. Execution goes through the Atlas Cloud skill, MCP or the command line.
Who it is for. For video makers, marketers and developers who generate clips with Seedance models through Atlas Cloud and want controllable results.
Good fit when
- You have a rough prompt for a clip and need it brought to a prompt with reference binding, camera and constraints
- You need a multi-shot sequence in which a character or product looks the same from shot to shot
- A clip must start and end on given frames: the skill builds a first-and-last-frame pair
- You need to extend an already generated clip or change one region in finished video
- You want to borrow phrasing from the library of ready prompts with previews
Not a fit when
- You need a different video model: for several models at once the repository has the Universal Video Prompt Skill
- You have no Atlas Cloud key and no generation budget: paid requests go through the provider's API
- You must get Seedance 2.5 specifically: the skill checks model availability and works on 2.0 until the provider has 2.5
Example request
Use seedance-2-5-skill: turn this idea into a 10-second clip about a coffee cup at dawn and show the storyboard firstLimitations
Generation needs an Atlas Cloud key and requests are paid. Seedance 2.5 is described as an upcoming launch, the default executable model is Seedance 2.0 for now, and the skill checks the provider's real capabilities. The 30 seconds, 4K and up to 50 references are, by the author's own note, only announced capabilities, and API parameters may differ. The repository license is undefined in GitHub metadata, while the badge in the README says CC BY 4.0.
How to disable. Delete the seedance-2-5-skill, universal-video-prompt-skill and atlas-cloud folders from the agent's skills directory. You can revoke the Atlas Cloud key in the account console.
FAQ
How do I install Seedance 2.5 Skill?
Run three commands: npx skills add AtlasCloudAI/awesome-seedance-2.5-prompts-skills --skill seedance-2-5-skill, the same with --skill universal-video-prompt-skill, and npx skills add AtlasCloudAI/atlas-cloud-skills --skill atlas-cloud. The third skill executes requests to Atlas Cloud and needs an API key.
Does the skill work with Seedance 2.5 right now?
Only if the selected provider exposes that model. By default the executable video model is Seedance 2.0, and storyboards come from Seedream 5.0 Pro.
Do I need to crop a storyboard before sending it?
By default no. If the panels stay readable, the whole storyboard goes in one reference request and Seedance reads the panel order itself. Make a clean copy without dividers and captions only if a test shows unwanted elements.
What should I do when a result is delayed?
Keep waiting on the same task id: the skill polls it every 2 seconds. A timeout or delay is no reason to send the paid request again.
What is in the prompt library?
More than a hundred ready prompts for Seedance. Each records a category, source platform, author, source link, input references and a preview video when one exists. The data lives in data/prompts.json and per-language files.
Discussion
No questions or reviews yet.
Related
SkillsA skill that teaches an agent to write prompts for ByteDance's Jimeng Seedance 2.0 video model: @ references, camera language, timing, templates
A dialogue that turns an idea into a storyboard and a prompt for the Seedance 2.0 video model in Jimeng, with six templates and ten examples
Awesome Seedance 2 Prompts
awesome-seedance-2-prompts
Over 2000 prompts for the Seedance 2.0 video model from ByteDance: cinema, anime, UGC, ads and memes, plus notes on the API and character consistency
The skill breaks a sales video into a storyboard and writes a prompt pack for the Seedance video model in Jimeng.
A skill that turns an idea, theme or novel excerpt into a 30-second multi-shot prompt for the Seedance 2.5 video model.
Two skills for generative video and images: video writes Seedance, Kling and Veo prompts by dramaturgy rules, image picks the model and its syntax
Comparison with similar tools
Cards that share the main search query, and how they differ.
| Tool | Official | Repository stars | Main difference |
|---|---|---|---|
| Seedance 2.5 Skill (this card) | no | 260 | |
| Seedance 2.0 Prompt Writing Skills | no | 4,413 | teaches writing prompts for Seedance 2.0 without running generation or storyboards |
| Kling CLI | yes | 1 | a skill for the Kling command-line tool that generates video and images |
README in short
The description combines a library of Seedance 2.5 prompts with the skill's installation. It explains the prompt structure (reference binding, action, space, camera, style, audio, constraints), the defaults and each prompt's provenance: category, platform, author and link. It cautions that a preview shows one observed output and no guarantee.
README badge
Original SKILL.md
--- name: seedance-2-5-skill description: >- Plan and generate controllable Seedance video using Seedream 5.0 Pro storyboards and Seedance 2.0 today, with a Seedance 2.5 route when available. Use for consistent people, products, objects, food, or scenes; storyboard-to- video; reference-to-video; first-and-last-frame image-to-video; extensions; and Atlas Cloud media generation. --- # Seedance 2.5 Skill ## Language route - For an English request, follow this file and the `*.md` references. - For a Chinese request, read [the Chinese workflow](references/workflow.zh-CN.md) first, then use the matching `*.zh-CN.md` reference files. - Keep model IDs, JSON keys, commands, media placeholders, and native-audio symbols exactly as code. Do not translate them. Do not force every request through one image-grid pipeline. Choose the video route first, then enable only the preparation modules the job needs. ## 1. Choose the creative route | Need | Route | Inputs | Result | |---|---|---|---| | One short, simple scene | T2V | Text prompt | One self-contained shot | | A multi-shot sequence with readable panels | R2V storyboard | One complete storyboard image | One request turns panel order into a continuous video | | A clip controlled by people, product, scene, or style assets | R2V asset references | A small role-specific asset pack | One clip built from those references | | A shot with an exact beginning and ending | I2V shot pair | Start and end keyframes | One independently reviewable shot | | An uninterrupted action exceeding the supported duration | Extend / chain | Previous generated tail frame and next prompt | Continuation of the same shot | | One continuous piece carrying several events | Staged whole-short | Text plus optional references | One request covering ordered stages, each landing on a stated end state | | A scoped change to an existing video | Editing | Source video plus target references | The source with only the named region or element changed | | A bridge between two finished clips | Seamless transition | Two videos | Generated bridge content between them | | Motion, blocking, and camera taken from a 3D preview | Blockout reference | Coarse or fine blockout video plus look references | Final render following the blockout's timing and staging | Use Seedance 2.0 as the executable default. Offer a Seedance 2.5 whole-short route only when the selected provider exposes that model and its actual limits. The last four routes depend on capabilities that differ per model and per provider. Verify availability before offering one; see [capabilities](references/capabilities.md) for the 2.0/2.5 comparison and for which published capabilities are platform features rather than API parameters. ### Storyboard or individual keyframes - Use the **whole storyboard image in one R2V request** by default whenever individual panels remain visually readable. Do not crop it first. Seedance can interpret the ordered panels as one continuous multi-shot video.