Generates videos with the Seedance 2.5 model using verbatim user prompts.
npx skills add https://github.com/DeepJH/doubao-skill-and-info --skill seedance-25-deepjh
Please help me install this Agent Skill. Skill: seedance-25 Source: https://github.com/DeepJH/doubao-skill-and-info/tree/main/skills/seedance-25 Command: npx skills add https://github.com/DeepJH/doubao-skill-and-info --skill seedance-25-deepjh
model_version to seedance_2.5 and refuses to downgrade or substitute other models. - Parameter Validation & Confirmation: Auto-suggests duration (5-30 seconds) and aspect ratio (16:9, 4:3, 1:1, 3:4, 9:16, 21:9) based on content type, then requires explicit user confirmation before generation. - Use Case: A marketer provides a finished prompt for a 15-second vertical product ad; the Skill confirms the exact prompt, 9:16 ratio, and duration, then generates the video without altering a single character. ## Quick Start Use the seedance-25 skill to generate a video with Seedance 2.5 using my exact prompt without any modifications.High-intent search queries and answers about installing and using this skill.
Provide your final prompt and state it must be passed verbatim. The Skill locks model_version to seedance_2.5, displays your exact prompt in a confirmation block, and only calls the video tool after your explicit approval.
Seedance 2.5 supports exactly six aspect ratios: 16:9, 4:3, 1:1, 3:4, 9:16, and 21:9. If you request an unsupported ratio, generation stops and you must choose from the supported list.
Seedance 2.5 accepts integer durations from 5 to 30 seconds. Requests outside this range are rejected with a prompt to pick a valid duration rather than switching to another model.
No. The Skill treats your prompt as an immutable string and forbids expansion, polishing, translation, or adding style keywords. Any changes require you to supply a new final version, which is then passed verbatim.
Generation stops and the Skill explicitly informs you that the model is unavailable. It never downgrades or substitutes a different model version on its own.
Retrieval triggers when prompts mention Doubao character imagery or contain time-sensitive words like latest, recent, or today involving brands, people, or events. Retrieved facts are listed separately in the confirmation and never merged into your prompt.
ChatCut-IncGenerate AI video clips from text, images, and reference footage.
hmzainjamilGenerate platform-specific captions, titles, hashtags, and scheduling recommendations for social media publishing.
4tieSearch the live web and fetch URL content as markdown.
lifegenieaiGenerate videos via Google Veo 3.1 API with custom prompts and parameters.