Agent skill

Higgsfield Character Design

by OSideMedia in OSideMedia/higgsfield-ai-prompt-skill

Pre-production story-and-character development for Higgsfield projects — the upstream layer that decides WHAT to prompt before any model runs.

MITAuto-check passedMedia & Creative

Install Higgsfield Character Design

skills CLI
$ npx skills add OSideMedia/higgsfield-ai-prompt-skill --skill higgsfield-character-design -a claude-code

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

GitHub CLI
$ gh skill install OSideMedia/higgsfield-ai-prompt-skill higgsfield-character-design --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/OSideMedia/higgsfield-ai-prompt-skill.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/higgsfield-character-design .claude/skills/higgsfield-character-design && 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
higgsfield-character-design
GitHub stars
713
Token cost
~6.4k tokens
SKILL.md length
3,507 words
Files
1
Skills in repo
33
Repo updated
First seen
Licence
MIT

At a glance

Pre-production story-and-character development for Higgsfield projects — the upstream layer that decides WHAT to prompt before any model runs.

  • Works in 6 steps: Premise → World Sheet (six dimensions) → Character (9 questions + the web) → …
  • The user wants to build a character
  • SKILL.md covers QUICK FACTS, The method: world first, Step 1 — Premise and Step 2 — World Sheet (six…, plus 10 more sections
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md

What it does

Higgsfield Character Design is an agent skill from OSideMedia/higgsfield-ai-prompt-skill. Pre-production story-and-character development for Higgsfield projects — the upstream layer that decides WHAT to prompt before any model runs. Use when the user wants to build a character, design a world, develop a story or premise, create a character sheet / character bible / story bible, lock a visual style or 'visual DNA', plan a multi-shot narrative with consistent characters, or says things like 'help me design a character', 'build the world', 'I need a backstory', 'make this character consistent across…

Its SKILL.md is about 6.4k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

It sits in Media & Creative, covering Comics and storyboards and Creative writing and fiction. The repository describes itself as: Claude AI skill for cinematic Higgsfield AI prompts — 32 sub-skills covering Seedance 2.5 (omni-reference, video edit + extend) and 2.0, the Hell Grind feature-film pipeline, an… The licence is MIT.

When your agent uses it

  • The user wants to build a character
  • Develop a story
  • Create a character sheet / character bible / story bible
  • Lock a visual style

Example prompts

  • “visual DNA”
  • “help me design a character”
  • “build the world”
  • “/higgsfield-character-design”

Workflow steps

6 steps, taken from the step headings in SKILL.md.

  1. Premise
  2. World Sheet (six dimensions)
  3. Character (9 questions + the web)
  4. Story Spine
  5. Style Sheet (Visual DNA + Forbidden List)
  6. Hand off to generation

What it can do on your machine

Read from SKILL.md and the folder at commit 7075497. 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

    No scripts in the folder and no shell commands in SKILL.md.

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

  • Network

    No URLs in SKILL.md.

    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

Higgsfield Character Design loads about 6.4k tokens when it runs. Until then it costs about 193 tokens; SKILL.md has 3,507 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~193
When it runs · the whole SKILL.md, loaded when a task matches
~6.4k

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 OSideMedia/higgsfield-ai-prompt-skill at commit 7075497, republished under its MIT licence (© OSideMedia). 3,507 words, ~6,435 tokens.

Download SKILL.mdSave it as .claude/skills/higgsfield-character-design/SKILL.md (or your agent's skills folder).
name
higgsfield-character-design
description
Pre-production story-and-character development for Higgsfield projects — the upstream layer that decides WHAT to prompt before any model runs. Use when the user wants to build a character, design a world, develop a story or premise, create a character sheet / character bible / story bible, lock a visual style or 'visual DNA', plan a multi-shot narrative with consistent characters, or says things like 'help me design a character', 'build the world', 'I need a backstory', 'make this character consistent across shots', 'develop my film/series concept', or 'I keep getting generic AI characters'. Routes the locked outputs into higgsfield-prompt + the right model. Adapted from Higgsfield's official character-design framework by @vavavinca.
user-invocable
true
metadata.tags
higgsfield, character-design, story, worldbuilding, character-sheet, story-bible, visual-dna, pre-production, consistency, narrative
metadata.version
1.3.0
metadata.updated
2026-09-26
metadata.parent
higgsfield

Higgsfield Character Design — Story Bible

The rest of this skill library answers "how do I prompt this?" This skill answers the question upstream of it: "what am I prompting, and why this and not generic AI slop?" It is the pre-production layer — premise, world, character, story, and a locked visual style — that you build before writing a single generation prompt, then hand off to higgsfield-prompt and the model guides.

Attribution. The framework, worksheet structure, and the "world first" method are adapted from Higgsfield's official Character Design materials by @vavavinca. We reuse the schemas and method; all examples here are original. Credit Vinca when sharing.

QUICK FACTS

Routing aids — read the linked sections for the actual method.

  • World first, always. Lock the world before you cast a character; the world is the gravity that shapes who they have to become →
  • Six steps, each locks an output you do not revisit: Premise → World → Character → Story Spine → Style Sheet → hand-off →
  • The flagship artifact is the 9-Question Character Sheet — thematic role, external/internal goal, psychological/moral need, wound, spark, silhouette, contradiction →
  • The strongest anti-slop tool is the Forbidden List in the Style Sheet — naming what the world is NOT is often more useful than the palette →
  • Specificity beats adjectives: if any character could say it, it's a stereotype; keep asking "why?" until the answer surprises you →
  • Fillable worksheets live in ../../templates/character-design/ — hand them to the user or fill them together →
  • Construction laws for the sheet as an artifact: plain grey background · creature sheets get two close-ups (mouth open + closed) · the face-lock crop · a size-ref frame for scale at extreme ratios no true landmark can express [FIELD] — grey and the one-face law are canonical in ../../templates/ad-asset-prep.md →
  • Once the character looks right, don't jump to scenes — run a screen test: casting read → role options → playable lines → voice triggers → one audition prompt [EMPIRICAL] →
  • This skill produces inputs; it does not generate. Hand the locked Visual DNA + character sheet to higgsfield-prompt →

The method: world first

A world is a system of constraints. Decide what the world will not let a character do, and you have decided who they have to become. So the order is not negotiable — each step locks an output the later steps depend on, and you don't loop back to relitigate it:

1. PREMISE      one arguable sentence, no characters yet
2. WORLD        six dimensions, filled before anyone is cast
3. CHARACTER    9 questions + a relationship web
4. STORY SPINE  10–20 beats, "therefore / but", never "and then"
5. STYLE SHEET  palette + lighting + materials + a FORBIDDEN list (the Visual DNA)
6. HAND OFF     feed the locked sheets into higgsfield-prompt + a model

The failure mode this prevents is the common one: open Higgsfield → generate a cool image → realize there's no story or consistency → quit. Build the bible first; the image is the receipt, not the recipe.

If the user arrives mid-stream ("I already have a character"), don't force them back to step 1 — locate where they are, check the upstream outputs exist (a character with no world is the usual gap), and fill the missing layer.


Step 1 — Premise

One sentence, no character names. It states an arguable claim about people or the world, not a plot. Four fields, then a pressure test (worksheet: ../../templates/character-design/premise.md).

  1. Theme statement — a claim, not a logline. "Belonging is something you build, not something you're owed."
  2. What it argues — the testable belief. "People who wait to be chosen stay invisible."
  3. The counter-argument — the opposing truth the story must honor, or the theme is propaganda. "Some people are kept out no matter what they build."
  4. Emotional promise — one word + one line. "Defiance — the thrill of a door forced open."

Pressure test (all five): compresses to ONE sentence (no "and", no semicolon) · contains NO character names · is arguable · has a real counter-argument · names a feeling, not just an event.


Step 2 — World Sheet (six dimensions)

Fill all six before casting anyone (worksheet: ../../templates/character-design/world-sheet.md). Each answer must be specific and must not change when characters arrive.

DimensionThe question it answersOriginal example
PhysicalLook, sound, smell, climate, geography, architecture"A tide-city built on stilts; the streets flood twice a day and everyone owns a boat before they own shoes."
SocialWho has power, how it passes"Status is measured in dry land. The few who own ground above the tideline never have to row."
EconomicCurrency, what's scarce, what people kill for"Fresh water is the real currency; salt ruins everything, so sealed jars are inheritance."
IdeologicalWhat's sacred, shameful, unspoken"Drowning is shameful — it means you didn't read the water. Nobody says a drowned name aloud."
HistoricalWhat everyone remembers or has forgotten"The Great Surge took the old capital a century ago; half the city pretends it's coming back."
SensoryThe daily textures the body knows"Rope-burned palms, the green smell of low tide, lantern oil, the constant give of a floor that moves."

Done check: read it aloud — does it sound like one world (no contradictions)? Is every answer specific? Will it stay true once characters walk in? Lock it. See also Hard vs soft worldbuilding for how knowable the rules should be.


Step 3 — Character (9 questions + the web)

The 9-Question Character Sheet

The framework's flagship artifact (worksheet: ../../templates/character-design/9-question-character-sheet.md). Header: name, age, world, archetype. Then, in order — vague answers are the enemy; only this character should be able to give them:

  1. Thematic role — what they prove or disprove about the premise. "That belonging can be built — he builds a crew out of other discards."
  2. External goal — a specific, currently-impossible want. "Buy the deed to a square meter of dry land before the next Surge season."
  3. Internal goal — what reaching it would mean about themselves. "Proof he's worth keeping, not just useful to keep around."
  4. Psychological need — the flaw that hurts THEM. "He treats every kindness as a debt he has to discharge before it's used against him."
  5. Moral need — the flaw that hurts OTHERS. "He recruits people for their use, then can't admit when he's done using them."
  6. Wound — the specific past event that taught them the world hurts. "At nine, watched his mother trade her water ration to a landlord who let her drown anyway."
  7. Spark — the specific past event that taught them what to become. "A stranger pulled him out of the tide and asked his name — the first person who wanted it."
  8. Silhouette — what their shape says across a room. "Always half-crouched at a doorway, weight on the back foot, ready to bolt or bargain."
  9. Contradiction — two things that don't match. "Pretends he travels light. Keeps a waterlogged ledger of everyone who ever did him a favor."
The Character Web

Characters reveal themselves in friction, never alone (worksheet: ../../templates/character-design/character-web.md). Map four spokes around the hero — Opponent (the primary opposite), Ally (who they lean on), Mentor (who taught them), Foil (the variation that exposes them). Characters can move through roles over the story (a mentor who becomes an opponent is a person; one stuck in a single bubble is a function). Diagnostic: if you can take one spoke away and the hero says the same line in every scene, the web is too thin.


Step 4 — Story Spine

10–20 numbered beats (worksheet: ../../templates/character-design/story-spine.md). Beat format:

Beat N: [character] [does / discovers / loses] [thing]. Result: [the next problem].

Connect beats with "therefore" or "but" — never "and then" (that's a list, not a story). Mark the Midpoint Turn: which beat flips the premise against the lead. Stop when the premise is proven — don't pad to hit 20.


Step 5 — Style Sheet (Visual DNA + Forbidden List)

Lock the look once, then inject it verbatim into every prompt (worksheet: ../../templates/character-design/style-sheet.md). Two ways to build it: (A) let the model draft it from the World Sheet (the worksheet carries a ready-to-paste prompt), or (B) fill it by hand. Six constants:

  • Palette — 5–7 hex codes (#1B3A4B, #6FA8A0, …). Inject the hex, don't describe colors in prose.
  • Lighting — the recurring light logic ("flat overcast noon, warm lantern pools after dark").
  • Materials — what the world is made of ("wet rope, salt-bleached wood, hammered tin, oilcloth").
  • Juxtaposition — one familiar + new pairing that fixes the world's identity ("a birthday party on a half-sunk rooftop").
  • Age & proportion — the build/era language for characters.
  • Real-life refs — 2–3 actors / paintings / films for the look (treat as look references, not identity to copy).

The Forbidden List is the most valuable field — name what the world is NOT. It's often more useful than the palette and it's your strongest anti-slop lever:

NO generic flooded-city teal-and-orange. NO neon. NO pristine surfaces — everything is salt-scarred. NO clear blue sky. NO photoreal celebrity faces. NO pure-black shadows; use deep indigo instead.

This Visual DNA — hex palette + forbidden list — is what gets pasted into prompts to keep a series consistent. See higgsfield-soul (Soul HEX) and higgsfield-style for how the color lock is applied at generation time.


Step 6 — Hand off to generation

This skill produces inputs; it does not generate. When the bible is locked, route forward:

  • The prompt → higgsfield-prompt (MCSLA structure). Inject the Visual DNA (hex + forbidden list) verbatim, and pull the subject from the character's Silhouette + Contradiction, the action from the relevant Story Spine beat.
  • The model → model-guide.md / image-models.md. For a character who recurs across many shots, train a Soul ID / Soul Cast identity (higgsfield-soul) rather than re-rolling one-offs; for a single hero image that won't reappear, a one-off generation is fine.
  • Multi-shot sequences → higgsfield-cinema (Cinema Studio) for shot-by-shot continuity; the Story Spine beats become the shot list.
  • Generic prompts get generic characters. A thin prompt ("a man's portrait, cyberpunk") cannot recover what the sheet would have supplied — the locked sheet is the difference between a function and a person on screen.
Ship the bible as a reusable artifact, not a paste

[EMPIRICAL — Joey story-bible-builder skill (2026-08-16), re-derived 2026-08-22] A bible that lives in a chat transcript gets re-explained every session, drifts a little each time it is retyped, and is the reason "the same" character comes back subtly different a week later. Write it out once as a single dense canon document and reuse the file — one artifact that every later prompt reads from, rather than context reassembled from memory.

Two ways it gets used, and they want slightly different shapes:

  1. Standalone canon reference — you consult it while writing prompts. Optimised for a human reading it.
  2. Context source for prompt work — it is loaded alongside the prompt so the world, characters, voices, tone and production rules are already known. Optimised for retrieval: short labelled sections, one fact per line, no narrative throat-clearing.

Keep it in workspace/input/ so it is found the way every other supplied document is (root ../../SKILL.md § Working Folders). For anything recurring, give each character a voice lock and a movement lock — one line each, fixed wording, reused verbatim — so speech register and physical signature stay pinned the way the Visual DNA pins the look. The locks fix the words, not where they go: the movement lock is copied verbatim wherever that movement can happen and transformed, not deleted, where it cannot (../higgsfield-acting/SKILL.md § Scene adaptation, ../shared/house-rulings.md P3-4); the voice lock goes into the audio field each time on Seedance 2.0 and into the role sentence on 2.5 — whether once per prompt or once per project is OPEN (../higgsfield-acting/SKILL.md § Voice, ../shared/house-rulings.md P2-2).

Build it by interview, not by questionnaire dump. Scope first (how big is this — one short, or a series?), then the spine (premise, thesis, timeline, aesthetic), then factions, locations and world rules, then characters, then relationships and ensemble dynamics, then the structural engines, then the production rules. Ask in that order and each answer constrains the next; ask everything at once and you get a pile of unrelated facts that contradict each other three sections later.

Only build the part that is needed. A single hero image does not need a faction table. Scope the interview to the work in front of you and leave the rest unbuilt — an unfinished bible is normal, an invented one is not.


Show full SKILL.md (1,570 more words)Show less

Sheet Construction Laws

[FIELD — AI-vs-VFX, 2026-08-08] The sections above decide what is on the sheet. These are the construction laws for the sheet as an artifact — the flaws they prevent are the ones you otherwise pay for in every downstream generation that references it.

Plain grey background, always. Not a set, not an environment, not a gradient. A sheet built on a busy background makes the downstream model decide which pixels are the character and which are the world, and that decision costs re-rolls. Grey is the credit-saver. Shade, the one-hex-per-project rule and every source's reason: ../../templates/ad-asset-prep.md § Design for win rate (the canonical home).

Creature and non-human sheets carry two head close-ups: mouth open and mouth closed. A creature detailed enough to be interesting is detailed enough to glitch between expressions — a jaw that reshapes the skull when it opens. Two close-ups at the same angle, one closed and one in a full roar, pin both extremes. The canonical layout is three panels: full body, head closed, head open.

The face-lock crop. A three-panel human sheet gives the model three faces to average — front full-body, back full-body, close-up — and at full-body scale the face is only a few dozen pixels, so averaging drags identity toward generic. Crop the heads out of the full-body panels. The full-body frames keep doing their real job (build, silhouette, wardrobe); the close-up becomes the single source of truth for the face. (The Hell Grind sheet removes only the front figure's head — both variants leave one readable face; the canonical statement is ../../templates/ad-asset-prep.md § Design for win rate.)

Scale between two subjects — a landmark sentence, a size-ref asset, or both. Relative size is the first thing to drift, and vague words do not fix it — "tiny compared to the enormous creature" returns a different size every generation, and worst on wides. A true, visible body landmark does hold: near-human props take a computed anchor sentence, and one feature anchored a thirty-metre giant (about 16× a human) with a landmark sentence and the human in frame (../higgsfield-seedance-2-5/VFX-PIPELINE.md § Stage 2, ../higgsfield-seedance/HELL-GRIND.md § Solutions born under deadline). An extreme ratio no true landmark can express needs its own asset; where both are possible, stack the landmark sentence and the size-ref (../shared/house-rulings.md P2-5). Build a size-ref frame: merge the two character sheets plus a frame whose scale was right into one image showing them together, write the proportion in human-height comparisons ("wingspan as wide as twenty humans lying head to toe"), save it as its own named asset, and attach it to every shot containing both. Add the asymmetry lock — if scale is uncertain, render the smaller subject smaller, never larger — because an undersized subject reads as distance while an oversized one reads as fake.

Fix a flawed sheet; don't rebuild it. A warped logo or a colour cast on an otherwise good sheet is a one-line edit on the model that holds inputs best (Nano Banana 2), not a re-prompt: change the logo to the one in image two. Re-prompting re-rolls everything that was already right. On the character's identity base, mask only the changed region back onto the untouched original (../higgsfield-soul/SKILL.md § The Untouched Base, ../shared/house-rulings.md P2-3). Model routing per asset class: ../../image-models.md § Routing by Asset Class.


Screen Test / Audition

[EMPIRICAL — community workflow] Adapted from a community casting-director system prompt. It works in the field; it is not a platform guarantee.

The bible tells you who the character is; the generated stills tell you they look right. Neither tells you how they move, speak, and react — and that is what every scene generation you are about to pay for depends on. So before spending on real scenes, run one cheap audition video and watch the character act. This is the same thesis as ../../templates/ad-asset-prep.md's generate-many → test-in-motion → lock-the-winner loop, applied to a person instead of a product: consistent-looking is not the same as castable.

Six steps, in order — the user chooses at step 5; never skip ahead to the final prompt:

  1. Casting read. Interpret the locked character sheet as casting material, not lore: presence, how weathered the character reads on screen, status, inner wound, external mask, likely voice, likely movement. The casting read is analysis for you; the audition prompt never states an age (../higgsfield-seedance/ENGINE-RULES.md rule 1 — describe by role, build and visible markers). Pull straight from the 9 questions — Silhouette and Contradiction do most of the work here.
  2. Role options (3–6). Offer contrasting role types the character could be cast as — lead, antagonist, mentor, tragic hero, villain-with-restraint, silent presence… Each option must state what the audition has to prove ("can he menace without raising his voice?"). An option with nothing to prove is not an option.
  3. Three audition lines. Real scene dialogue with subtext, implying an off-camera reader, each playable more than one way. Not trailer narration, not lore recitation — if it sounds like a voice-over, it cannot be acted.
  4. Three voice triggers. Each is ≤3 comma-separated qualities ("authoritative, grief, controlled") — a how, never the spoken line itself. The three must contrast with each other, not be synonyms.
  5. The user picks one role + one line + one trigger (blending is allowed — the blend still obeys ≤3 qualities).
  6. One final audition prompt, under 3500 characters. Build a playable scene, not a showcase:
    • Simple setting — the audition room, not the movie: one location, no spectacle competing with the face.
    • Camera: MCU or locked-off eye-level, shallow depth of field. The frame holds still so the performance is what moves.
    • Acting spine: objective, obstacle, tactic, emotional shift, and the subtext under the line.
    • Reaction beats and silence — write the listening, not just the speaking.
    • At least one expressive vocal moment (a stutter, a vocal crack, a whispered threat, a controlled shout) and subtle physical acting (blinks, gaze shifts, jaw tension) — muscle, not emotion labels.
    • Unresolved ending — cut the take before the moment closes; resolution reads as trailer, open tension reads as acting.
    • Finish with film grain + realistic skin, so you judge a performance, not a render.

Divergence rule: if the user comes back and picks a different role type, rebuild the acting logic from scratch — new objective, new tactics, new beat pattern. A mentor read is not a villain read with softer adjectives; the same beats reworded is the audition-room version of slop.

Worked mini-example (the tide-city fixer from Step 3)
  • Casting read: reads more worn than his build suggests; low status wearing borrowed confidence; mask of the indispensable broker over a wound of being kept only while useful; voice quick and transactional, dropping register when cornered; moves like his Silhouette — half-crouched at doorways, weight on the back foot.
  • Role option — villain-with-restraint. Must prove: he can threaten someone while doing them a favor, without ever raising his voice.
  • Audition line: "You've still got both your water jars. That's not luck — that's me. So sit down, and let me tell you the second half of the favor." (Playable warm or menacing; the off-camera reader is whoever owes him.)
  • Voice trigger: quiet, transactional, barely controlled.
Direct the takes, then lock the winner
  • Facial direction: for muscle-level control of the audition takes (the jaw tension, the asymmetric almost-smile), hand the beats to ../higgsfield-facs/SKILL.md — AU codes, not emotion labels.
  • After the audition: when a take proves the character is castable, lock that winner's identity with ../higgsfield-soul/SKILL.md (Soul ID / Soul Cast) so every real scene starts from the version that can act — the lock-the-winner step of the ../../templates/ad-asset-prep.md loop.

Anti-generic drills

Three fast tests for killing slop, usable at any step:

  • Always Ask Why. Repeat "why?" until the answer surprises you. "He's guarded." Why? "He was betrayed." Why does that still run him? "Because the one who betrayed him also taught him everything." A stereotype is any line any character could say; yours is the line only this one could.
  • One Wound, Two Characters. The same wound produces opposite people. "Abandoned as a child" → one runs toward every relationship and clings; another refuses to need anyone. Pick the less obvious reaction.
  • Eyes Wide Shut / Silhouette test. Can you recognize the character from shape alone, with no face? If the silhouette doesn't read (warrior / mystic / rogue), the design is generic — fix it before generating.

Hard vs soft worldbuilding

How knowable should the world's rules be? Decide with three questions (worksheet note in world-sheet.md):

  1. Does the story depend on the audience mastering the rules? Yes → hard (rules are explicit and consistent, e.g. heist/system stories).
  2. Do you want the audience feeling smart or feeling wonder? Smart → hard; wonder → soft.
  3. Are you willing to leave things unexplained? Yes → soft (mystery preserved, e.g. fable/dream logic).

Hard worlds reward consistency in your prompts (repeat the same physical rules); soft worlds reward atmosphere over explanation. This choice changes how literal your World Sheet's Physical/Historical dimensions should be at generation time.


Templates

Fillable worksheets in ../../templates/character-design/ (hand them to the user or fill collaboratively):

WorksheetStep
premise.md1 — theme + counter-argument + pressure test
world-sheet.md2 — six dimensions + hard/soft check
9-question-character-sheet.md3 — the flagship character schema
character-web.md3 — Opponent / Ally / Mentor / Foil map
story-spine.md4 — beats + midpoint turn
style-sheet.md5 — Visual DNA + Forbidden List (+ AI-draft prompt)

  • higgsfield-prompt — turns the locked sheet into an MCSLA generation prompt (the primary hand-off)
  • higgsfield-soul — Soul ID / Soul Cast for recurring-character consistency; Soul HEX for the palette lock
  • higgsfield-cinema — Cinema Studio multi-shot sequences; Story Spine → shot list
  • higgsfield-style — applying the Visual DNA / forbidden list at generation time
  • model-guide.md · image-models.md — picking the model for the character/scene

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

Files

Just SKILL.md in skills/higgsfield-character-design of OSideMedia/higgsfield-ai-prompt-skill.

Open the folder on GitHubat commit 7075497

Compare with similar skills

Higgsfield Character Design 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.

Higgsfield Character Design compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Higgsfield Character Design this skillOSideMedia/higgsfield-ai-prompt-skill713—~6.4kAutomated safety check: PassMIT
Novel Storyboardeternityspring/shuohao-skills4.3k—~1.8kAutomated safety check: NotesApache-2.0
Storyboard Managerailabs-393/ai-labs-claude-skills4551 repos~4.6kAutomated safety check: PassMIT
Novel Storyboardsuihe1/short-drama-production181—~2.1kAutomated safety check: NotesApache-2.0
Adaptationdanjdewhurst/story-skills2861 repos~3.5kAutomated safety check: NotesMIT
Film Crewguaardvark/guaardvark257—~727Automated safety check: PassMIT

Similar skills

  • Novel Storyboard

    eternityspring/shuohao-skills

    给 AI 短剧出分镜:三层结构——段(一次视频生成,≤15 秒)→ 分镜(段内 2–5 秒的剪切,认领剧本节拍) → 分镜图(每切一张关键帧:主分镜图钉 0.00 秒,子分镜图钉各自切点)。

    4.3k GitHub stars~1.8k tokensUpdated 2 days ago
    Media & CreativeAuto-check: notes
  • Storyboard Manager

    ailabs-393/ai-labs-claude-skills

    Assist writers with story planning, character development, plot structuring, chapter writing, timeline tracking, and consistency checking.

    455 GitHub starsUsed in 1 repo~4.6k tokens
    Media & CreativeAuto-check passed
  • Novel Storyboard

    suihe1/short-drama-production

    将剧本与导演意图转成生成段、逐镜时间、参考素材计划和可校验分镜报告;使用官方 h3-prompt-writing 编写 H3 提示词,可导出投产包。用于技术分镜、镜头表和分镜修订;生成视频须另有明确任务与成本授权,不负责最终选片或剪辑。

    181 GitHub stars~2.1k tokensUpdated 25 days ago
    Media & CreativeAuto-check: notes
  • Adaptation

    danjdewhurst/story-skills

    This skill should be used when the user asks to "make an audiobook", "narration script", "narrator", "ACX", "Findaway", "pronunciation guide", "how long is the audiobook", "adapt to a screenplay"…

    286 GitHub starsUsed in 1 repo~3.5k tokens
    Media & CreativeAuto-check: notes
  • Film Crew

    guaardvark/guaardvark

    Run Guaardvark's five-role Film Crew locally: screenwriter, casting (Cast Library LoRAs), cinematographer storyboards, keyframe renders, editor.

    257 GitHub stars~727 tokensUpdated yesterday
    Media & CreativeAuto-check passed
  • Drives long-form fiction, scripts, storyboards, interactive films and long-document translation through InkOS, with every change made by a typed action.

    10k GitHub starsUsed in 1 repo~1.1k tokens
    Writing & ContentAuto-check passed

More from OSideMedia/higgsfield-ai-prompt-skill

All 33 skills in this repo
  • Higgsfield

    OSideMedia/higgsfield-ai-prompt-skill

    A skill your agent uses whenever the user asks anything about Higgsfield AI — writing or refining video/image prompts, choosing a model (Kling, Veo, Wan, Seedance, Minimax Hailuo, DoP, Soul, Nano…

    713 GitHub stars~9.1k tokensUpdated 14 days ago
    Auto-check passed
  • Higgsfield Assist

    OSideMedia/higgsfield-ai-prompt-skill

    A skill your agent uses when the user asks about Higgsfield Assist (the built-in GPT-5 copilot), how to use the platform's native AI assistant, credit optimization strategies, plan selection, how to…

    713 GitHub stars~2.9k tokensUpdated 14 days ago
    Auto-check passed
  • Higgsfield Image Shots

    OSideMedia/higgsfield-ai-prompt-skill

    A skill your agent uses when the user wants to generate a cinematic still image on Higgsfield, asks about shot framing, camera angle, or composition for image prompts, needs a specific shot type…

    713 GitHub stars~5.1k tokensUpdated 14 days ago
    Auto-check passed
  • Higgsfield Mixed Media

    OSideMedia/higgsfield-ai-prompt-skill

    A skill your agent uses when the user asks about Mixed Media, wants to apply artistic preset styles to an image (Noir, Sketch, Paper, Canvas, Particles, Neon, etc.), combine multiple artistic…

    713 GitHub stars~2.3k tokensUpdated 14 days ago
    Auto-check passed
  • Higgsfield Moodboard

    OSideMedia/higgsfield-ai-prompt-skill

    A skill your agent uses when the user asks about Moodboard, building a moodboard from reference images, curated moodboard presets, Soul Hex color transfer, applying a visual style direction to…

    713 GitHub stars~2.3k tokensUpdated 14 days ago
    Auto-check passed
  • Higgsfield Motion

    OSideMedia/higgsfield-ai-prompt-skill

    A skill your agent uses when the user wants to apply a named Higgsfield motion preset, asks about VFX presets, transformation effects, elemental effects, or transition presets.

    713 GitHub stars~5k tokensUpdated 14 days ago
    Auto-check passed

Questions about Higgsfield Character Design

What does Higgsfield Character Design do?

Pre-production story-and-character development for Higgsfield projects — the upstream layer that decides WHAT to prompt before any model runs. Higgsfield Character Design is an agent skill from OSideMedia/higgsfield-ai-prompt-skill. Pre-production story-and-character development for Higgsfield projects — the upstream layer that decides WHAT to prompt before any model runs.

When should I use Higgsfield Character Design?

Higgsfield Character Design fits situations like: the user wants to build a character; develop a story; create a character sheet / character bible / story bible; lock a visual style.

How do I install Higgsfield Character Design in Claude Code?

Run `npx skills add OSideMedia/higgsfield-ai-prompt-skill --skill higgsfield-character-design -a claude-code`. Or copy the skill folder (skills/higgsfield-character-design in OSideMedia/higgsfield-ai-prompt-skill) into .claude/skills/higgsfield-character-design in your project. Claude Code loads it when a task matches its description.

How do I install Higgsfield Character Design in Codex?

Run `npx skills add OSideMedia/higgsfield-ai-prompt-skill --skill higgsfield-character-design -a codex`. Or copy the skill folder (skills/higgsfield-character-design in OSideMedia/higgsfield-ai-prompt-skill) into .agents/skills/higgsfield-character-design in your project. Codex loads it when a task matches its description.

Can I use Higgsfield Character Design 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 OSideMedia/higgsfield-ai-prompt-skill --skill higgsfield-character-design -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/higgsfield-character-design, .gemini/skills/higgsfield-character-design, .github/skills/higgsfield-character-design and .opencode/skills/higgsfield-character-design in your project.

What does Higgsfield Character Design need to run?

SKILL.md names no scripts, command-line tools or credentials: Higgsfield Character Design is instructions for the agent only.

Does Higgsfield Character Design access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Higgsfield Character Design 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 Higgsfield Character Design use?

Higgsfield Character Design is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Higgsfield Character Design use?

About 6.4k tokens (SKILL.md is roughly 26k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Higgsfield Character Design?

Skills that share tags, products or a category with Higgsfield Character Design: Novel Storyboard (eternityspring/shuohao-skills, 4.3k stars), Storyboard Manager (ailabs-393/ai-labs-claude-skills, 455 stars), Novel Storyboard (suihe1/short-drama-production, 181 stars) and Adaptation (danjdewhurst/story-skills, 286 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Higgsfield Character Design?

OSideMedia (a GitHub user) maintains it in OSideMedia/higgsfield-ai-prompt-skill, which has 713 GitHub stars. The repository holds 33 skills in this directory. The repository was last updated on September 27, 2026.

Source: OSideMedia/higgsfield-ai-prompt-skill on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.