Generate Higgsfield cinematic prompt drafts with camera-move presets and Soul aesthetics.
npx skills add https://github.com/antonyfmunoz/OS --skill higgsfield-antonyfmunoz
Please help me install this Agent Skill. Skill: higgsfield Source: https://github.com/antonyfmunoz/OS/tree/main/skills/tools/higgsfield Command: npx skills add https://github.com/antonyfmunoz/OS --skill higgsfield-antonyfmunoz
This Skill turns messy, generic “make a video” requests into precise Higgsfield-ready prompts that preserve subject identity and deliver intentional cinematic camera moves instead of blurry or mushy results.
Use the higgsfield skill to draft a Higgsfield DoP image-to-video request for a 5s “Crash Zoom In” hook, including the subject anchor, one beat of action, and a short list of style tokens suitable for Antony to run through the web GUI.
High-intent search queries and answers about installing and using this skill.
To write image-to-video prompts with specific cinematic camera moves, map your brief to a single camera-move preset like Crash Zoom In or Bullet Time, and use short single-beat prompts with action and style tokens to lock motion.
A single-beat prompt for AI video generation is a short instruction containing one action and style tokens. This format locks motion and prevents over-specification, ensuring the model avoids conflicting camera directions.
To prevent identity drift in image-to-video AI generation, apply Soul aesthetic presets for a consistent look and optionally include Soul ID or reference frames. This maintains subject identity consistency throughout the camera-move sequence.
Yes, you can use webhook workflows for headless AI video generation by utilizing REST API submit/poll or webhook callbacks. This supports webhook-driven output handling with request_id-based idempotency for EOS-ready production patterns.
Prompt engineering for cinematic video works without conflicting camera directions by aligning all camera language with a chosen camera-move preset and restricting instructions to one beat of action per prompt.
m2ai-portfolioConvert user concepts into cinematic Seedance 2.0 prompts using the FRAMES framework.
OpenRouterTeamSubmit, poll, and download videos from OpenRouter's asynchronous generation API.
tuanqt98Generate two A/B short-form video scripts with hooks, CTAs, and timestamped scenes.
OSideMediaGenerate structured Higgsfield AI prompts using the MCSLA formula.
