agentsclimarketplace

Higgsfield prompting

Skill produtoramaxvision/maxvision-utilities/media-forge/skills/higgsfield-prompting

MaxVision's curated Claude Code marketplace — production-grade skills for n8n workflows, Google Tag Manager API, and more.

Install
npx -y skills add produtoramaxvision/maxvision-utilities --skill higgsfield-prompting

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

  • no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
  • 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Higgsfield prompting playbook — MCSLA formula, DoP camera verbs, Cinema Studio lens dictionary, Soul ID lifecycle, Marketing Studio templates. Trigger when caller plans a Higgsfield generation.

SKILL.md

5.4 KB, ~1.4k tokens by cl100k_base, as published. Nobody here has run it

Higgsfield Prompting Playbook

1. MCSLA Formula (universal)

Every Higgsfield prompt benefits from these five concentric layers, in order:

LayerQuestionExample
M MotionWhat moves?"the bartender slides a glass across the counter"
C CameraHow does the lens move?"crane_up + dolly_in" (DoP verbs)
S SubjectWho/what is the focal point?"a noir-era bartender in mid-thirties, scar under left eye"
L LightingWhat kind of light, from where?"single tungsten bulb above, hard rim from window left"
A AestheticWhat is the emotional / style register?"1940s Hollywood noir, high contrast, slight bloom"

Compose in this order; the platform's Soul models tokenise Motion + Camera first, then resolve Subject inside the implied Lighting + Aesthetic frame.

2. DoP Camera Verb Cheatsheet (22 verbs)

VerbUse when
dolly_inPush closer to subject (heighten intimacy / suspense)
dolly_outPull back to reveal context
crane_upRise above scene (god-view, reveal)
crane_downDescend onto subject (introduce / land into action)
orbitCircle subject (showcase 360°, build awe)
crash_zoomSnap zoom-in (reactive, comedic, jolt)
bullet_timeFrozen-orbit (Matrix; iconic)
fpv_droneFirst-person drone weave (action / sports)
handheldNaturalistic, slight wobble (documentary / vérité)
whip_panFast horizontal sweep (energy transition)
tilt_up / tilt_downVertical reveal
pan_left / pan_rightHorizontal reveal
arcArc move (parallax + reveal)
truckSideways tracking (parallel to action)
pedestalVertical lift without tilt (architectural)
rack_focusShift focal plane between foreground / background
vertigo_effectDolly-zoom (Hitchcock — anxiety / dread)
staticNo camera motion (composition is the statement)
low_angle / high_angleStatic angle modifiers

Best practice: combine ≤2 verbs per shot. More than 2 collapses into incoherent motion.

3. Cinema Studio 3.5 Lens Dictionary

FieldTypical ValuesEffect
focalLengthMm14, 24, 35, 50, 85, 135Wide → tele (compression)
apertureFStop1.4, 1.8, 2.0, 2.8, 5.6, 8Shallow → deep DOF
sensorSizefull-frame, super35, apsc, m43, imaxSensor crop + look
colorGradingteal-orange, bleach-bypass, noir, pastel, vibrant, plus free-formEra / mood
lensIde.g. arri-master-prime-35mm, cooke-s4-50mm, zeiss-master-anamorphic-50mmSpecific lens character

Recipe for "Sundance indie drama":

{ "focalLengthMm": 35, "apertureFStop": 2.0, "sensorSize": "super35", "colorGrading": "bleach-bypass" }

Recipe for "Wes Anderson symmetry":

{ "focalLengthMm": 35, "apertureFStop": 8, "sensorSize": "full-frame", "colorGrading": "pastel", "lensId": "cooke-s4-35mm" }

4. Soul ID Lifecycle Best Practices

  1. Train once per character — pick 3-7 source images covering frontal, 3/4, profile, varied expression. Training is 250 credits ($9.75 on Plus).
  2. Name aggressively — character name is the lookup key. Use unique distinguishing names: not "Sarah" but "Sarah Nguyen, lead reporter".
  3. Mark used after every successful generationmedia_higgsfield_soul_id markUsed updates the LRU.
  4. Audit periodicallymedia_higgsfield_soul_id list shows training cost + last_used; archive IDs unused >90 days.
  5. One Soul ID per project character — do NOT train multiple IDs for the same character; consistency degrades.

5. Marketing Studio Template Decision Tree

Caller intentTemplateBest for
Show product in uselifestyleCasual / aspirational
Open the box, reveal itemunboxingE-commerce, hype
Hard-sell with VOtv-spotDTC brand launches
Showcase texture/materialasmrBeauty, food, fabric
Lateral motion + product flipshyper-motionSneakers, gadgets, drinks
Customer talks to cameratestimonialTrust building
Influencer-style POVugcSocial commerce
Pros vs cons walkthroughproduct-reviewTech, comparison
Short looped beatsreelReels, TikTok, Shorts

Always include a clean productUrl — the platform crawls product imagery from it.

6. Mode → Tool Map

Mode requestedUse tool
Pure t2v with aesthetic presetmedia_higgsfield_soul_id (to fetch ID) + provider generate
i2v with camera motionmedia_higgsfield_dop
Cinematic lens controlmedia_higgsfield_cinema_studio
Talking head (photo + audio)media_higgsfield_speak
Product UGCmedia_higgsfield_marketing_studio
Swap character in existing videomedia_higgsfield_recast
Score a candidatemedia_higgsfield_virality_predictor

Gives 0 of the 12 instructions most prompt engineering skills give in ~1.4k tokens

Counted across 563 of the 626 authors here whose files we hold, read 2026-08-06

  • ask at most three clarifying questionsin 22 of 563, across 15 files
  • respond in the user input languagein 14 of 563, across 9 files
  • preserve the original intentin 13 of 563, across 11 files
  • Establish baseline metrics and collect representative examplesin 12 of 563, across 2 files
  • Identify failure modes and prioritize high-impact fixesin 12 of 563, across 2 files
  • Apply prompt and workflow improvements with measurable goalsin 12 of 563, across 2 files
  • Roll back quickly if quality or safety metrics regressin 12 of 563, across 2 files
  • validate changes with tests and roll out in controlled stagesin 12 of 563, across 2 files
  • generate quantitative baseline performance reportsin 12 of 563, across 2 files
  • create representative test scenariosin 12 of 563, across 2 files
  • treat prompts as codein 12 of 563, across 5 files
  • test prompts on diverse inputsin 12 of 563, across 8 files

Said here and by no other author read

  • Order prompt layers as motion, camera, subject, lighting, then aesthetic
  • Combine no more than two camera verbs per shot
  • Use the lens dictionary for cinematic controls
  • Train exactly one Soul ID per project character
  • Use three to seven source images to train a Soul ID
  • Give each Soul ID a unique distinguishing name

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.