with one click
image-blast-3d
// Generate one specified atomic 3D object. Use when the user names exactly one object instance to make, or provides one image plus the object name/description.
// Generate one specified atomic 3D object. Use when the user names exactly one object instance to make, or provides one image plus the object name/description.
| name | image-blast-3d |
| description | Generate one specified atomic 3D object. Use when the user names exactly one object instance to make, or provides one image plus the object name/description. |
Create exactly one 3D object for project $0.
The arguments must identify one object clearly enough for this forked skill to work alone.
$0 is missing, ask for the world slug.ls -a before reading generated state.worlds/$0/output/<object>/object.json, worlds/$0/image.json, or worlds/$0/source/*.json.object.json, create the minimal durable intent:{
"schema_version": 1,
"world": "$0",
"object": {
"id": "<object-slug>",
"name": "<object name>",
"description": "<literal object description>",
"materials": [],
"source_images": [],
"evidence": [],
"generate_as_3d_object": true,
"working_dir": "worlds/$0/output/<object-slug>"
},
"updated_at": "..."
}
description, materials, source_images, and evidence.--image-edit-prompt. Use this as the base prompt, filling in the target object and then elaborating with object-specific details: "Isolate the from this image. Reproduce it exactly as shown -- same colors, materials, and proportions. White background, centered, tight crop, studio lighting. No other objects, no scene, no people, no text, no shadows on the ground. Isolate the object and remove all clustered, adjacent, overlapping, or items resting on the target object. Create a clean render of that one single object that is true to the source image."object.json.Run the generator and wait for it to finish:
node .claude/scripts/asset-pipeline/generate-single-asset.mjs --world "$0" --object-id "<object-id>" --image-edit-prompt "<object-specific extraction prompt>"
The first run may create a clean reference PNG with image edit. Later runs reuse the latest reference PNG by default, including --regenerate model runs. Use --regenerate-reference only when the user asks for a new extraction from the source image. For iterative refinement, pass --reference-only to stop after the image-edit step.
If request metadata records provider URLs but local model or image files are missing, fill them from the matching hidden request JSON:
node .claude/scripts/project/ensure-local-assets.mjs --from "<request-json-path>"
Hunyuan is the default 3D provider. Hunyuan defaults are --face-count 50000, --enable-pbr true, and --generate-type Normal. If the user asks for more detail, polygon reduction, or a white geometry-only model, pass the matching options:
--face-count <40000-1500000>--generate-type Normal|LowPoly|Geometry--polygon-type triangle|quadrilateral for LowPoly--enable-pbr true|falsePass --provider meshy only when the user asks for Meshy. Meshy defaults are:
{
"topology": "triangle",
"target_polycount": 30000,
"symmetry_mode": "auto",
"should_remesh": true,
"should_texture": true,
"rigging_height_meters": 1.7,
"animation_action_id": 12,
"enable_safety_checker": true,
"enable_animation": false,
"enable_rigging": false,
"enable_pbr": true
}
For Meshy-specific requests, pass the matching options:
--target-polycount <integer>--topology triangle--symmetry-mode auto--should-remesh true|false--should-texture true|false--enable-animation true|false--enable-rigging true|falseFor explicit model regeneration from the existing reference, append --regenerate. For a new source extraction and model, append --regenerate-reference. For direct single-image generation, use:
node .claude/scripts/asset-pipeline/generate-single-asset.mjs --world "$0" --image "<image-path>" --object-name "<object-name>" --description "<description>" --image-edit-prompt "<object-specific extraction prompt>"
Final response: report the object id, output directory, generated model files, and any failed/resumable request metadata.
Create, inspect, and manage an IMAGE-BLAST project envelope under worlds/<slug>. Use before other image-blast skills (image-blast-uncover, image-blast-world, image-blast-3d, etc.) or whenever the user asks about active project state.
Main image analysis skill, object generation, plate decisionmaking, and initial scene description. Use this skill when user wants to create objects and necessary for main image analysis.
Generate the 3D static environment of a world from a source or plate image.
Generate one clean plate/source cleanup image for an IMAGE-BLASTER world. Use after confirmed objects exist, when removing foreground objects or other specified content from a source image.
Generate one image edit from explicit input images and a prompt. Use for source cleanup, clean plates, object removal, or other FAL-backed image edits.
Generate and post-process a clean sound effect, object impact, or ambient loop.