agentsclimarketplace

02 3d cgi

Skill pixelab-ch/higgsfield-skills/skills/02-3d-cgi

Model-aware Claude Agent Skills for AI image & video generation on Higgsfield — 15 creative skills with opt-in MCP generation. Reworked from beshuaxian/higgsfield-seedance2-jineng.

Install
npx -y skills add pixelab-ch/higgsfield-skills --skill 02-3d-cgi

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

One thing to look at

  • 1 stars1 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

Crafts 3D CGI and rendered video prompts for Higgsfield and routes to the best model (seedance_2_0 primary, wan2_7 fallback). Use when the user wants 3D rendered, CGI, Pixar-style, Unreal Engine, photorealistic 3D, computer- generated, or digitally rendered video. Triggers on: 3D animation, CGI, rendered, Blender, Unreal Engine, octane render, ray tracing, volumetric, subsurface scattering, physically based rendering, or any 3D/CG video request.

SKILL.md

6.1 KB, as published. Nobody here has run it

Higgsfield 3D CGI Skill

What this skill does

Crafts production-ready 3D CGI and rendered video prompts and routes them to the best Higgsfield model for computer-generated output. On request, it assembles the full parameter set, surfaces the credit cost, and — after explicit user confirmation — generates the asset via the opt-in generation flow.

The difference between a generic "make it look 3D" request and a production-ready CGI prompt lies in specificity: render engine terminology, material properties, lighting language, camera dynamics, and the architectural "hooks" that grab attention in the first two seconds.


Model routing

Primary and fallback video models

seedance_2_0 (primary)wan2_7 (fallback)
RationaleBroadest param set for CGI: genre tags, 21:9 aspect, full media-role supportUse only when primary is unavailable
Resolution480p, 720p, 1080p (def 720p)720p, 1080p
Aspect ratiosauto, 21:9, 16:9, 4:3, 1:1, 3:4, 9:1616:9, 9:16, 1:1, 4:3, 3:4no 21:9
Duration4–15 s (continuous range)2–15 s (continuous range)
Tunable paramsresolution, mode {std,fast}, genre {auto,action,horror,comedy,noir,drama,epic}None beyond resolution and aspect
Media rolesimage, start_image, end_image, video, audiostart_image, end_image, audio only

Routing rule: Use seedance_2_0 by default. Only fall back to wan2_7 when (1) the primary model is unavailable, or (2) the user explicitly accepts wan2_7's constraints after being told what they give up: no 21:9 aspect ratio, no genre or mode params, and no image or video media roles. If the user has not been informed of these trade-offs, stay on seedance_2_0 and let them decide — never switch silently.

MODEL-06 directive: If a parameter is rejected at generation time, call higgsfield:models_explore with the target model name to re-verify the live schema. Full parameter tables: references/model-specs.md.

Per-platform quick rule

Vertical 9:16 for TikTok / Reels / Shorts; 16:9 for YouTube and LinkedIn; 21:9 cinematic widescreen only on seedance_2_0. Full per-platform table: references/model-specs.md.


Prompt-building workflow

  1. Gather intent — Confirm with the user: CGI style (Pixar, photorealistic, low-poly, cel-shaded, isometric, etc.), subject, mood, target platform, aspect ratio, desired duration, and any I2V reference images.

  2. Select model — Apply the routing table above. Choose seedance_2_0 by default; select wan2_7 only when a routing trigger applies.

  3. Build the prompt — Use the craft principles in the references below:

  4. Present for review — Show the assembled prompt and all parameters to the user for review and refinement before any generation call. The user must approve the prompt text, model, aspect ratio, duration, and (if used) genre/mode.


Opt-in generation

Generation costs Higgsfield credits and requires explicit user confirmation before any generate call is made. This skill never auto-generates.

Full step-by-step flow (confirmation gate, balance/cost surface, generate → poll → display): ../../shared/generation-flow.md

This skill's primary model: seedance_2_0

Media upload: Conditional — only when the user supplies a start_image, end_image, video, or audio reference. This skill is NOT a mandatory-I2V skill; T2V is the default path. When the user provides a reference file, run the higgsfield:media_uploadhiggsfield:media_confirm atomic pair before generating.

Tool signatures: ../../shared/mcp-tools.md


Reference materials

FileContents
references/model-specs.mdPer-model parameter tables, aspect ratios, durations, routing rationale, and re-verify directive for seedance_2_0 and wan2_7
references/render-craft.mdRender style encyclopedia (9+ styles), material science library (20+ surfaces), lighting setups (15+), camera movement guide (15+), particle and effects language
references/hooks.md2-second hook framework with 12+ hook techniques and prompt phrasing templates for CGI
references/examples.mdMaster prompt template, 5 full production-quality CGI example prompts (Pixar, photorealistic product, isometric sci-fi, creature transformation, architectural flythrough), and common-mistakes guide

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.