TypingMindTypingMindThe best frontend for AI
Higgsfield Repurpose logo

Higgsfield Repurpose

Community
OSideMedia
higgsfield-repurpose

Use when the user already has a FINISHED video and wants Higgsfield to repurpose, restyle, clip or analyze it: Shorts Studio (restyle one 4–120s source video into a set of short-form clips from a style preset), Clipify (turn one YouTube video into subtitled clips), Virality Predictor (predict virality, hook strength, attention, retention risk with a dashboard), or Video Analysis (scene-by-scene breakdown of an uploaded video or a YouTube link). Also use when the user asks which of these cost credits and which are free, 'turn my video into shorts', 'will this go viral', 'score my ad', 'break this video down scene by scene'. Names: shorts_studio_create, shorts_studio_create_preset, shorts_studio_list_presets, shorts_studio_status, clipify, virality_predictor, brain_activity, video_analysis_create, video_analysis_status.

Overview

PublisherOSideMedia
Repositoryhiggsfield-ai-prompt-skill
Skill namehiggsfield-repurpose
Stars
722
Forks
126
Bundled files
Instructions only
LicenseMIT
Links
Skill sourceRepositoryHomepage
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by OSideMedia on GitHub. Read the source before you install it.

Installation

Install the Higgsfield Repurpose AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/OSideMedia/higgsfield-ai-prompt-skill/tree/main/skills/higgsfield-repurpose
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/OSideMedia/higgsfield-ai-prompt-skill.git /tmp/higgsfield-ai-prompt-skill
mkdir -p .claude/skills
cp -r /tmp/higgsfield-ai-prompt-skill/skills/higgsfield-repurpose .claude/skills/higgsfield-repurpose
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Higgsfield Repurpose in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Higgsfield Repurpose on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Higgsfield Repurpose is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Higgsfield Repurpose — Shorts, Clips, and Analysis of Finished Video

QUICK FACTS

Routing aids — read the linked sections for the full rules. Nothing in this sub-skill is field-tested.

  • Four surfaces take a video that already exists: Shorts Studio (restyle → shorts), Clipify (YouTube → subtitled clips), Virality Predictor (a score), Video Analysis (a scene breakdown) →
  • Paid vs free is per tool — shorts_studio_create is PAID; its get_cost estimate and preset creation state they are free; list / status calls state no cost →
  • Shorts Studio: source 4–120s, output 720p only, 9:16 default (16:9 optional); preflight with get_cost: true + duration_seconds — no preset or upload needed for the estimate →
  • Shorts Studio presets store a STYLE (≤10 media, videos ≤30s, public https URLs) — write the preset prompt as laws, not adjectives →
  • Clipify: exactly ONE YouTube URL per job; 1–20 clips; 9:16 / 1:1 / 16:9; subtitle font, case, position and highlight colour are parameters →
  • Virality Predictor: input is a confirmed uploaded video or a completed generated video; it is a PREDICTION — compare variants with it, don't treat it as audience data →
  • Video Analysis: exactly one of uploaded video / YouTube URL; ~3–5 min; warn up front: the longer the video, the less accurate the breakdown →
  • Burned-in captions, cuts, soundtrack, voice swaps are other surfaces — hand-off map →

What this sub-skill is for

Every surface here starts from a video that already exists — a finished ad, a long YouTube video, a generated clip — and produces something derived from it: restyled shorts, cut-down clips, a performance prediction, or a scene list. None of them writes a generation prompt from scratch; that is the rest of this library.

SurfaceInputOutputTool / model
Shorts StudioOne uploaded video, 4–120s + a style presetA set of AI-restyled short-form clipsshorts_studio_* (MCP)
ClipifyOne YouTube URLReady-to-share clips with subtitlesclipify (video catalog model)
Virality PredictorAn uploaded or generated videoVirality / engagement / attention / hook / retention-risk prediction + dashboardvirality_predictor (MCP)
Video AnalysisAn uploaded video or a YouTube URLScene-by-scene analysisvideo_analysis_* (MCP)

Not this sub-skill: changing what is inside a clip (swap an object, re-light, extend it) — that is ../higgsfield-seedance-2-5/SKILL.md (video_edit / video_extension), ../higgsfield-seedance-vfx/SKILL.md, or Genjutsu (../higgsfield-marketing-studio/SKILL.md § 14). Many edited versions of one ad → Ad Multiplier, same § 14.

Paid or free — read this before calling anything

[OFFICIAL — Higgsfield MCP tool schema, 2026-09-26] unless marked.

ToolSpends credits?Evidence
shorts_studio_createPAID — "reserves credits"Tool description
shorts_studio_create with get_cost: true + duration_secondsFree — "estimate the credit cost without submitting a job"; no preset or source video neededTool description
shorts_studio_create_presetFree — "just stores a STYLE — no generation, no credits"Tool description
shorts_studio_list_presets · shorts_studio_list_sessions · shorts_studio_statusNot stated — list / status calls that submit nothing; not called free hereTool descriptions (silent on cost)
clipifyNot stated. It is a video-catalog model, so it runs through generate_video / the CLI and can be preflighted with get_cost: true (MCP) or higgsfield generate cost clipify … (CLI — confirm the flag form with higgsfield model get clipify first)[OFFICIAL — platform, 2026-09-26] snapshot has no price field
virality_predictorNot stated, and the tool has no get_cost fieldTool schema
video_analysis_createNot stated, and the tool has no get_cost fieldTool schema
video_analysis_status · video_analysis_jobsNot stated — status / list calls that submit nothing; not called free hereTool descriptions (silent on cost)

Where the schema is silent, do not call it free. Check the balance before and after (balance / transactions on MCP, higgsfield account status on the CLI) the first time a surface is used, and record what it actually cost. No prices appear in this file because none were measured.

Shorts Studio

[OFFICIAL — Higgsfield MCP tool schema, shorts_studio_*, 2026-09-26]

"Start a Shorts Studio short: restyle one uploaded source video (4s–120s) into a set of AI-generated short-form clips using a style preset. PAID — reserves credits."

ParamValuesNotes
source_video_iduuidThe uploaded video's video_input id (media_upload_widget, type=video). Required unless get_cost
preset_iduuidFrom shorts_studio_list_presets or shorts_studio_create_preset. Required unless get_cost
preset_sourceuser · cmsThe preset item's own preset_source. Required unless get_cost
aspect_ratio9:16 (default) · 16:9Output orientation
resolution720pThe only value — the schema pins it as a constant
get_costboolWith duration_seconds, returns the estimate and submits nothing
duration_seconds> 0, ≤ 120Source length; required when get_cost is true

The run, in order:

  1. Estimate first — get_cost: true + duration_seconds (free; no preset, no upload).
  2. Pick or make a style — shorts_studio_list_presets returns the user's own presets first, then the CMS library (paginated via next_cursor); or make one (below).
  3. Upload the source — the user uploads through media_upload_widget (type=video); the tool asks for an upload before it is called, never a chat attachment.
  4. Create — PAID. Returns a session with empty job_ids.
  5. Poll shorts_studio_status until clips appear. status='completed' means every clip job is terminal, not necessarily successful — check each clip.

Schema mismatch to know about: shorts_studio_status says to poll each clip's job_id "via job_status", but no job_status tool is in the connector's 2026-09-26 tool list — the listed job tools are jobs_wait, job_display and show_generation_by_ids. Which one the status tool means is unverified.

Not published: how many clips a session produces, how long each is, and what the restyle keeps from the source (timing, audio, framing). Look at one real session before promising any of them.

Style presets (free)

shorts_studio_create_preset stores a style from reference media — no generation, no credits.

  • Limits: ≤ 10 media total; each video ≤ 30s (send duration so the cap applies); every media item a public https URL (use an uploaded media's URL, or media_import_url first).
  • Fields: name (required), prompt (optional style direction), image_medias, video_medias, thumbnail.
  • Naming: Higgsfield's own tool instruction is to give an unnamed preset a random friendly two-word name rather than ask — follow it.
  • What this library adds [INFERENCE — untested]: a style preset is a saved style block, the same job as Cinema Studio's Manual Style — so write its prompt the same way: laws, not descriptions (grade behaviour, lighting law, texture, camera register), no vague labels, no director name-drops. See ../higgsfield-cinema/SKILL.md § Manual Style — Authoring Guide.

Clipify

[OFFICIAL — platform, 2026-09-26] — model clipify, "Turn one YouTube video into ready-to-share clips with subtitles" (video catalog, ../../specs/models_explore_snapshot_2026-09-26.json).

ParamValuesDefault
urlsYouTube URL array — "Provide exactly one URL; submit one Clipify job per source video"required
clips_num1–2010
clip_aspect9:16 · 1:1 · 16:99:16
subtitle_fontnotosans · notoserif · notosansdisplay · ibmplexsans · mplusrounded1c · bebasneue · archivoblack · unbounded · inter · montserrat · bangers · permanentmarker · playfairdisplay · caveatnotosans
subtitle_caseupper · lower · as-isas-is
subtitle_positionbottom · center · topbottom
subtitle_highlight_hex#RRGGBB#FFE84D
track_face_cropbool — "Track faces when cropping clips"true
max_height144–2160 — "Maximum source processing height"1080
segment_seconds2–60 — "Segment duration in seconds"10

Clipify vs Shorts Studio. Clipify cuts one YouTube video into subtitled clips; it has no style or restyle parameter. Shorts Studio restyles one uploaded video into AI-generated clips from a preset. Long talk / podcast / tutorial → clips: Clipify. A short piece whose look should change: Shorts Studio.

The source is fetched from YouTube — clip only videos the user owns or has the rights to. What segment_seconds controls beyond its one-line description is not documented.

Virality Predictor

[OFFICIAL — Higgsfield MCP tool schema, virality_predictor, 2026-09-26]

"predicts a video's virality potential, engagement, attention, audience response, retention risk, hook strength, and creative performance with an interactive dashboard."

  • action: "create" with params.medias: [{role: "video", id: <uuid>}] — the id is a confirmed uploaded video media_id or a completed generated video job_id; params.model is the constant "virality_predictor".
  • action: "preview" with params.job_id re-opens an existing dashboard.
  • CLI face: Higgsfield's own skills repo says its Virality Predictor path "Uses Virality Predictor (brain_activity) with --video" [OFFICIAL — Higgsfield skills repo README, v0.12.0]; CLI 1.1.23 model get brain_activity shows a text-output model named "Brain Activity" that requires medias. That the two are the same product rests on that README line — not verified here.

How this library uses it [INFERENCE — untested]:

  • It is a prediction, not audience data. This repo has never compared its scores with real performance. Report it as "the predictor says", never as "this will go viral".
  • Compare variants, don't chase a number. The useful question is relative: hook A vs hook B on the same ad, cut 1 vs cut 2. Rewriting a creative only to move one score is optimizing the instrument.
  • Natural pairings: hook variants from ../higgsfield-marketing-studio/SKILL.md, Ad Multiplier versions (same skill, § 14), campaign batches in ../higgsfield-content-factory/SKILL.md.

Video Analysis

[OFFICIAL — Higgsfield MCP tool schema, video_analysis_*, 2026-09-26]

  • Input — exactly one of: video_input_id (an uploaded video's media id) or youtube_url (https on youtube.com, www.youtube.com, m.youtube.com or youtu.be only).
  • Queued: returns status='queued'; poll video_analysis_status every 30–60 seconds until completed (scenes populated) or failed (fail_reason populated). "Processing typically takes 3-5 minutes on average." Past jobs: video_analysis_jobs.
  • The warning the tool requires, verbatim: "warn the user up front that the longer the video, the less accurate the scene-by-scene analysis becomes — short clips give the most reliable results." Say it before submitting, every time.
  • Practical consequence: trim to the sequence you actually care about before uploading.

Uses in this repo [HYPOTHESIS — UNMEASURED]:

  • Reference breakdown — a short reference sequence → a scene list to seed ../higgsfield-shotlist-director/SKILL.md or to audit with ../higgsfield-scene-engine/SKILL.md. Analyze a reference for its structure; the shotlist still gets written, not copied.
  • Cut check — did a generated multi-shot clip actually cut where the prompt declared ("strictly N shots")? The analysis is a second pair of eyes, not a verdict.
  • The output's shape beyond "scenes" is not documented. Treat what comes back as notes to verify against the video, never as ground truth.

Where the neighbouring jobs live

The user wantsSurfaceWhere it is covered
Captions burned into the pixelsHiggsfield's subtitles workflow../higgsfield-stack/SKILL.md § Higgsfield's bundled workflows
Cuts, trims, soundtrack, layouts, animated text, title cards on footageHiggsfield's video-editing workflow (Higgsedit)same
The same video at another aspect ratioreframe (MCP tool; also a CLI workflow — verify params with higgsfield workflow get reframe --json)—
A different voice on the same videovoice_change../higgsfield-audio/SKILL.md § Voice change and voice cloning
One object swapped, or motion transferredGenjutsu../higgsfield-marketing-studio/SKILL.md § 14
Many independently edited versions of one adAd Multipliersame

Provenance

  • Shorts Studio, Virality Predictor, Video Analysis: [OFFICIAL — Higgsfield MCP tool schema, 2026-09-26] — read from the connector's tool definitions; no tool was called.
  • Clipify: [OFFICIAL — platform, 2026-09-26] — the video models_explore snapshot.
  • brain_activity: CLI 1.1.23 model get (a free schema read) plus Higgsfield's skills repo README v0.12.0.
  • Nothing here has been run; costs, clip counts, clip lengths and the accuracy of any analysis are unmeasured.

Related skills

  • ../higgsfield-stack/SKILL.md — execution surfaces, preflight, Higgsfield's bundled workflows
  • ../higgsfield-audio/SKILL.md — voice change and cloned voices
  • ../higgsfield-marketing-studio/SKILL.md — ad creative, Ad Multiplier, Genjutsu
  • ../higgsfield-content-factory/SKILL.md — campaign batches that virality scores can rank
  • ../higgsfield-shotlist-director/SKILL.md · ../higgsfield-scene-engine/SKILL.md — where a scene breakdown goes next

Frequently asked questions

What does the Higgsfield Repurpose AI skill do?

Use when the user already has a FINISHED video and wants Higgsfield to repurpose, restyle, clip or analyze it: Shorts Studio (restyle one 4–120s source video into a set of short-form clips from a style preset), Clipify (turn one YouTube video into subtitled clips), Virality Predictor (predict virality, hook strength, attention, retention risk with a dashboard), or Video Analysis (scene-by-scene breakdown of an uploaded video or a YouTube link). Also use when the user asks which of these cost credits and which are free, 'turn my video into shorts', 'will this go viral', 'score my ad', 'break...

Why use Higgsfield Repurpose on TypingMind?

Because you install it once and use it with any model. Higgsfield Repurpose is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Higgsfield Repurpose in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/OSideMedia/higgsfield-ai-prompt-skill/tree/main/skills/higgsfield-repurpose. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Higgsfield Repurpose?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Higgsfield Repurpose?

As many as you like. As long as a model supports skills, you can use Higgsfield Repurpose with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Higgsfield Repurpose AI skill free?

Yes. It is published on GitHub by OSideMedia under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

Related AI Skills

View all
nexu-io logo

Hyperframes

nexu-io
OrganizationPopular

Create video compositions, animations, title cards, overlays, captions, voiceovers, audio-reactive visuals, and scene transitions in HyperFrames HTML. Use when asked to build any HTML-based video…

11.5K
98.9K
40 files
View
nexu-io logo

Audio Jingle

nexu-io
OrganizationPopular

Audio generation skill — jingles, beds, voiceover, and sound effects. Routes music requests to Suno V5 / Udio / Lyria, speech to MiniMax TTS / FishAudio / ElevenLabs V3, and SFX to ElevenLabs SFX or…

11.2K
96.7K
1 files
View
openclaw logo

Proof Video

openclaw
OrganizationPopular

Add subtitles, captions, narration cues, or zoom to a proof video or PR recording using repo-local capture helpers and a system ffmpeg renderer.

82.2K
391K
Instructions
View
openclaw logo

Facetime

openclaw
OrganizationPopular

Inspect FaceTime internal stages, place an approved call to a configured owner handle, and request hangup through facetime_call.

82.2K
391K
1 files
View

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇