Scenario 3D Asset Workflows
Overview
Generate a concept image first, then run image-to-3D: it has a larger catalog and more art direction
than text-to-3D (txt23d). Per-family contracts: scenario-meshy, scenario-rodin,
scenario-sparc3d. Walkable scenes and Gaussian splats: scenario-3d-worlds. PBR retexture of a
finished mesh or scene: scenario-patina-retexture. Inside Blender, through the Scenario for
Blender extension: scenario-blender. Connection and the core loop: the scenario skill. If a
sibling skill named here is missing from your available skills, load it with skill_get (Scenario
MCP) or ask the user to install it (https://github.com/scenario-labs/skills#install); unattended,
proceed from tool schemas and flag the gap.
Quick reference
| Step |
Tool |
Notes |
| Find 3D models |
recommend with the capability and the user's own words (search for a name) |
Capabilities: txt23d, img23d, 3d23d |
| Inspect inputs |
model_schema_get |
Always call before model_run |
| Generate |
model_run |
Pass reference images as asset IDs |
| Wait |
jobs_wait |
Any job_id returned without assets goes in job_ids; never poll job_get in a loop |
| Preview |
asset_display |
Interactive GLB/FBX/VOX/OBJ viewer on MCP App hosts |
| Download |
asset_download |
Returns a URL; save with curl -L |
Worked example: game-ready treasure chest
- Plan check. Before step 1, run
recommend for every later step (step 2, plus any rig or
retexture). An entry marked requires_plan_upgrade fails with a 403
(ModelAccessRestrictedError) only after earlier steps were paid: settle the upgrade or pick
another member first. If the img23d call answers ask_user about existing concepts, take the
concept-image branch: step 1 makes them.
- Batch price. Price per asset from one real concept:
dry_run each candidate on it (later legs
per step 4), times the count. At authoring time one LLM-based member quoted 1,300 CU at its
default effort, a single-image Tripo generator 45. Past the plan's credits, each job's wait grows
with its cost.
Generate the concept: pick a text-to-image model via recommend with the user's own words as
prompt, then model_schema_get and model_run. Prompt a single centered subject on a plain
background. Show a non-biped creature at eye level, from the side or three-quarters, every leg
visible and clear of the body: top-down hides leg undersides and body height. A biped headed for
a rig takes the pose in Rigging and animation. Upload a user's reference with upload_asset and
pass its asset ID instead. Keep a reference on a plain background as is, and prefer it to a busy
shot: removal keeps a glossy subject's reflections of the scene, which image-to-3D bakes into the
texture. Remove a busy one's background first (recommend the removal) and pass that output's id
instead. An isolated subject matters, a transparent cutout does not; an image field's
background_note, when present, says the same.
recommend with capability="img23d" and the user's own words as prompt. Live members include
the Hunyuan 3D, Meshy, Tripo, and Trellis families, and Scenario's own LLM-based 3D generation:
GPT-6 Astra 3D and Claude Opus 5.5 3D (the cheaper at authoring time; compare both with
dry_run). Those two LLM-based members alone:
- reconstruct one object from 1 to 8 photos or renders into an editable mesh with named parts and
per-part PBR materials;
- suit props and hard-surface subjects, not characters or creatures;
- price by
buildEffort: standard, high (the default), maximum;
- cap the exported triangle count with
faceBudget (the game range in step 3);
- change a result by editing it (see Refining meshes).
model_schema_get: inspect image inputs, polycount, PBR, and topology controls. Target 10k to
50k triangles for this prop; never assume defaults meet that budget. Use only exposed controls:
- Numeric field: read
min and max, or the description when they are absent.
- Enum: read
allowed_values and their topology and count meanings (scenario-rodin covers
qualityMeshOption). A quad count is not a triangle count.
- No face control, or the target is below its range: generate, then remesh down if needed
(Refining meshes). Never invent a face field;
scenario-sparc3d covers members without one.
Check the output's triangle count against the budget. At authoring time:
- Tripo (
faceLimit) and Meshy (targetPolycount, set where scenario-meshy says) reached both
ends. At 10k, an eight-legged creature kept every leg and claw, and a chest its padlock and
handles. Meshy's Smart Topology members capped lower, at 15k or 30k.
- Hunyuan 3D (
faceCount, from 40k) covers only the top of the range, and Trellis
(decimationTarget, from 100k) starts above it. Remesh only for a target under the member's
floor.
model_run with parameters={"image": "asset_xxx", ...} and dry_run=true to price it. Then
run it for real with wait=false and pass its job_id to jobs_wait (job_ids=["<job_id>"])
until no pending_job_ids remain. A downstream step (rig, retexture) has no payload to dry_run
until its input mesh exists: quote it from recommend, then re-price with dry_run on the real
asset before launching.
asset_display the output asset_id, then asset_download and curl -L -o chest.glb "<url>".
Omit format when asset_get shows properties.format glb: it returns the stored original
byte for byte. format converts to glb, fbx, or obj; pass glb for the LLM-based members,
which store a Blender file. Ask for fbx for a Unity project without a glTF importer, and keep
the GLB too: at authoring time the FBX embedded only two images, so it may miss PBR maps.
Prompt-only build. Both LLM-based members support txt23d. When the user wants no concept to
approve, skip step 1. In step 2, recommend with capability="txt23d" and the user's words as
prompt. Follow steps 3 to 5 with images: []; describe the object and name required parts in
prompt. A chest's lid, body, and padlock returned as those node names.
Multi-view models take several images of one subject; read the count and fields off
model_schema_get. Side fields follow the subject: on the Tripo multi-view members at authoring
time, leftImage was the character's own left (nose toward frame left), rightImage the mirror. A
sheet labeled by the viewer's side reverses them, and a swapped pair can fuse front and back into
one shell.
Texture and lighting controls
Several image-to-3D families split texture from geometry. Read each dial in model_schema_get
before promising a look. Authoring-time examples from the Tripo members:
texture: false returns a bare mesh and is a cost_impact field: the cheap path when the user
textures in a DCC. textureQuality also moves the price; an empty texture version picker keeps
the provider's default.
delight strips lighting baked into the reference so the mesh relights in the engine. It
defaulted to on; turn it off only when painted shading is the art style. The export stays a lit
material either way: for no relighting at all, the user sets an unlit shader on import.
pbr: the P2 generators defaulted it off (on forces texturing, billed as such); the 3.1
generators and the texturing member defaulted it on, then ignored the texture parameters. Read its
description and say which won.
- No Tripo member exposed symmetry: a result unlike the vendor's own app may come from an app-only
setting.
- Geometry and texture take separate seeds (
seed, textureSeed): hold seed and the image, vary
textureSeed for texture variants on one shape.
autoSize scales the output to real-world meters and defaulted to off: set true for a
real-scale import. dimensions still reads the unit-normalized mesh (longest side 1.0), since the
meters sit in the GLB's node transform: not a failure.
Inspecting results
Hosts without MCP Apps get the app_url link instead of the viewer. The viewer's capture button
calls capture_3d_view, an app-only tool that posts an image asset_id of the current view into
the conversation. Reuse it as a generation reference or a similarity search image. Never call
capture_3d_view yourself: it needs PNG canvas data only the viewer has.
Refining meshes
3D-to-3D utilities retexture, remesh, UV-unwrap, and segment parts: recommend with
capability="3d23d" plus the operation in the user's own words. Most take the source asset_id in
a kind: "3d" file field, usually model (also mesh, file3d, sceneFile). A 400
Input model is required or Provide a reference image or a 3D model means the mesh went in under
another name (an image field, or a URL).
The LLM-based members of step 2 also edit an existing mesh, prop or character:
sceneFile takes its asset_id (any 3D model, returned in the same format), images stays
empty, prompt states the change.
buildEffort fits the change, not the object: standard for a recolor, high or maximum to
add or reshape parts.
- Price follows
buildEffort, not the mesh, and recommend priced no edit at authoring time: a
dry_run on any project mesh quotes it.
- A recolor kept every part name and the face count. A reshape is uncapped (
faceBudget ignored):
it took a 4,817-triangle cat to 788,954, so plan a remesh before an engine.
- A replaced material keeps its old name; only added parts get new ones. The edit is generated:
check what the prompt did not name against the source.
- GPT-6 Astra 3D's description also claimed Mixamo-named rigs, ARKit face shapes, animation loops,
UV re-unwraps, and modular kits, so it competes in those
recommend calls.
Splitting a finished mesh into parts is a contested lane, so it stays a recommend pick. At
authoring time vendors offered granularity controls or a bundled PBR retexture, and image-to-parts
members built parts from the picture instead. A convincing mesh is not a game-ready one: part
separation, joints, materials, and animation are each their own pass. Polygon caps differ by
topology: read the slider's max instead of promising a count. A named provider feature is a valid
search target, but an empty result or one member's enum does not prove a capability absent
platform-wide. recommend the unmet need (a hinged prop), inspect the returned schemas, and report
it unverified if none exposes it. Never invent a call from the provider's site.
A staged chain toward game-ready
One image-to-3D run bakes its texture onto UVs the generator chose. A mesh headed for an engine
usually goes further, each output's asset_id feeding the next tool's kind: "3d" field:
- Geometry: a multi-view member when several angles exist, with
texture: false where offered, and
pbr: false beside it where pbr overrides the texture settings.
- UV unwrap.
- Retexture the unwrapped mesh from a prompt, an image, or a style reference.
- Rig last, on the topology that ships.
Each step after the first is a 3d23d recommend pick and its own billed run: dry_run it on the
real input before launching. Along the chain:
- Retexture schemas. Avoid a setting that rebuilds UVs, which discards the unwrap. Check the
same
pbr default (the Tripo texturing member defaulted to PBR at authoring time). When a schema
is silent on UVs, prefer a member with an explicit keep-existing-UVs setting.
- Input caps. One unwrapper took meshes up to 30,000 faces at authoring time: set the
generator's face limit under it (the Tripo multi-view member exposed
faceLimit and
smartLowPoly), or remesh down first.
- Checks between steps.
asset_display each output. Without a viewer, compare face count, UVs,
and dimensions from asset_get step to step, plus the thumbnail asset, which can lag the job by
seconds. hasUVs proves UVs exist, not a clean layout; an unwrapper's "albedo" PNG may be a
preview, not the atlas.
- Inherited defects. Faces rendering dark or inside out are flipped normals, which later steps
inherit. Scale and pivot can shift between steps, even within one vendor (
autoSize above):
confirm both in a DCC before the rig.
- Sibling formats. Feed the GLB onward unless the next schema asks otherwise or the mesh is
quads, which stay in FBX. With no GLB, feed the OBJ or FBX the next schema lists (a
dry_run may
skip the format check, so the real run confirms it), else convert with asset_download format.
Fixing normals, origin, or naming by hand is DCC work, not an MCP call. The
scenario-blender-expert skill drives Blender; it installs locally, so skill_get cannot fetch it.
Rigging and animation
Rigging is a separate 3d23d step on a finished mesh, not a generator flag: recommend that
capability plus the rigging need in the user's own words.
Body plan picks the model. Humanoid riggers take the mesh plus at most a front-facing hint or an
approximate height, and infer a biped skeleton. Non-bipeds go to a model exposing rigType:
quadruped, hexapod, octopod, avian, serpentine, aquatic. A static prop has no body plan:
if recommend answers ask_user, part segmentation is the nearer step.
A biped headed for a rig starts from a T- or A-pose concept (full body, front view, arms clear of
the torso, plain background): riggers infer joints from the pose. A non-biped keeps step 1's view.
For quads:
- Pick an image-to-3D member whose schema offers quad topology and set it; the default is triangles.
At authoring time the Tripo P2 and 3.1 members had a
quad switch with a lower face cap;
scenario-rodin and scenario-meshy name their families' controls.
- Quad output arrives as FBX. glTF stores triangles only, so
format: "glb" or a GLB-only step
triangulates it.
- Keep the FBX, rig it with a member whose file field lists FBX, and check each output's face type
in a DCC before promising quads. A rig returned as GLB is triangulated: deliver the quad FBX
beside it and say so.
Three schema details decide whether the output is usable:
- Formats. Riggers accept GLB, often OBJ, FBX, or STL, and expose no output-format field. The
rig comes back as GLB or FBX, mostly unannounced: expect a DCC pass when the engine needs the
other.
- Size ceiling. A file
max_size is rare (one humanoid rigger capped at 30 MB): check the
schema before decimating.
- Animation versus rig. Where a rigger exposes
animation, it retargets a preset clip; its
allowed_values are rig-type prefixed (quadruped:walk), so read them. By default only the
retarget file comes back: set includeRiggedModel to keep the plain rig too. The two arrive as
separate assets with identical metadata, in no set order. asset_get tells them apart: the
retarget has properties.hasAnimations true and an animationFrameCount, the plain rig false and
null.
When only motion is wanted, motion-transfer video models animate a still character image with no
skeleton: see scenario-video. Video-to-motion models that auto-rig an uploaded mesh are the one
place an outputFormat enum picks the engine target.
Common mistakes
- Running
model_run without model_schema_get: 3D parameters differ far more between models than
image parameters do.
- Hardcoding model IDs: catalogs rotate (a
deprecated:<replacement_id> tag names the successor).
Re-discover each session: recommend for a need, search for a name.
- Promising a named skeleton, an influence count, or bones for wings and extra limbs the chosen
member does not document: pick the closest body plan, then finish in a DCC.
- Sending a biped to a
rigType model: the enum has no biped value; humanoids have their own
riggers.
1---2name: scenario-3d3description: Use when generating or handling 3D assets through the Scenario MCP server, including text-to-3D or image-to-3D meshes, GPT-6 Astra 3D or Claude Opus 5.5 3D, GLB, FBX, OBJ, STL, or VOX files, PBR-textured or game-ready models, voxel models, multi-view reconstruction, editing a mesh by prompt, retexture, remesh, UV unwrap, auto-rigging a biped or quadruped, or retargeting an animation, previewing a mesh in the 3D viewer, or downloading a model for import into Unity, Unreal, Godot, or Blender.4license: MIT5metadata:6 source: https://skillmd.com/skills/scenario-labs/scenario-3d7---89# Scenario 3D Asset Workflows1011## Overview1213Generate a concept image first, then run image-to-3D: it has a larger catalog and more art direction14than text-to-3D (`txt23d`). Per-family contracts: `scenario-meshy`, `scenario-rodin`,15`scenario-sparc3d`. Walkable scenes and Gaussian splats: `scenario-3d-worlds`. PBR retexture of a16finished mesh or scene: `scenario-patina-retexture`. Inside Blender, through the Scenario for17Blender extension: `scenario-blender`. Connection and the core loop: the `scenario` skill. If a18sibling skill named here is missing from your available skills, load it with `skill_get` (Scenario19MCP) or ask the user to install it (https://github.com/scenario-labs/skills#install); unattended,20proceed from tool schemas and flag the gap.2122## Quick reference2324| Step | Tool | Notes |25| -------------- | ------------------------------------------------------------------------------ | -------------------------------------------------------------------------------------- |26| Find 3D models | `recommend` with the capability and the user's own words (`search` for a name) | Capabilities: `txt23d`, `img23d`, `3d23d` |27| Inspect inputs | `model_schema_get` | Always call before `model_run` |28| Generate | `model_run` | Pass reference images as asset IDs |29| Wait | `jobs_wait` | Any `job_id` returned without assets goes in `job_ids`; never poll `job_get` in a loop |30| Preview | `asset_display` | Interactive GLB/FBX/VOX/OBJ viewer on MCP App hosts |31| Download | `asset_download` | Returns a URL; save with `curl -L` |3233## Worked example: game-ready treasure chest3435- **Plan check.** Before step 1, run `recommend` for every later step (step 2, plus any rig or36 retexture). An entry marked `requires_plan_upgrade` fails with a 40337 (`ModelAccessRestrictedError`) only after earlier steps were paid: settle the upgrade or pick38 another member first. If the `img23d` call answers `ask_user` about existing concepts, take the39 concept-image branch: step 1 makes them.40- **Batch price.** Price per asset from one real concept: `dry_run` each candidate on it (later legs41 per step 4), times the count. At authoring time one LLM-based member quoted 1,300 CU at its42 default effort, a single-image Tripo generator 45. Past the plan's credits, each job's wait grows43 with its cost.44451. Generate the concept: pick a text-to-image model via `recommend` with the user's own words as46 `prompt`, then `model_schema_get` and `model_run`. Prompt a single centered subject on a plain47 background. Show a non-biped creature at eye level, from the side or three-quarters, every leg48 visible and clear of the body: top-down hides leg undersides and body height. A biped headed for49 a rig takes the pose in Rigging and animation. Upload a user's reference with `upload_asset` and50 pass its asset ID instead. Keep a reference on a plain background as is, and prefer it to a busy51 shot: removal keeps a glossy subject's reflections of the scene, which image-to-3D bakes into the52 texture. Remove a busy one's background first (`recommend` the removal) and pass that output's id53 instead. An isolated subject matters, a transparent cutout does not; an image field's54 `background_note`, when present, says the same.552. `recommend` with `capability="img23d"` and the user's own words as `prompt`. Live members include56 the Hunyuan 3D, Meshy, Tripo, and Trellis families, and Scenario's own LLM-based 3D generation:57 GPT-6 Astra 3D and Claude Opus 5.5 3D (the cheaper at authoring time; compare both with58 `dry_run`). Those two LLM-based members alone:59 - reconstruct one object from 1 to 8 photos or renders into an editable mesh with named parts and60 per-part PBR materials;61 - suit props and hard-surface subjects, not characters or creatures;62 - price by `buildEffort`: `standard`, `high` (the default), `maximum`;63 - cap the exported triangle count with `faceBudget` (the game range in step 3);64 - change a result by editing it (see Refining meshes).653. `model_schema_get`: inspect image inputs, polycount, PBR, and topology controls. Target 10k to66 50k triangles for this prop; never assume defaults meet that budget. Use only exposed controls:67 - **Numeric field:** read `min` and `max`, or the description when they are absent.68 - **Enum:** read `allowed_values` and their topology and count meanings (`scenario-rodin` covers69 `qualityMeshOption`). A quad count is not a triangle count.70 - **No face control, or the target is below its range:** generate, then remesh down if needed71 (Refining meshes). Never invent a face field; `scenario-sparc3d` covers members without one.7273 Check the output's triangle count against the budget. At authoring time:74 - Tripo (`faceLimit`) and Meshy (`targetPolycount`, set where `scenario-meshy` says) reached both75 ends. At 10k, an eight-legged creature kept every leg and claw, and a chest its padlock and76 handles. Meshy's Smart Topology members capped lower, at 15k or 30k.77 - Hunyuan 3D (`faceCount`, from 40k) covers only the top of the range, and Trellis78 (`decimationTarget`, from 100k) starts above it. Remesh only for a target under the member's79 floor.80814. `model_run` with `parameters={"image": "asset_xxx", ...}` and `dry_run=true` to price it. Then82 run it for real with `wait=false` and pass its `job_id` to `jobs_wait` (`job_ids=["<job_id>"]`)83 until no `pending_job_ids` remain. A downstream step (rig, retexture) has no payload to `dry_run`84 until its input mesh exists: quote it from `recommend`, then re-price with `dry_run` on the real85 asset before launching.865. `asset_display` the output `asset_id`, then `asset_download` and `curl -L -o chest.glb "<url>"`.87 Omit `format` when `asset_get` shows `properties.format` `glb`: it returns the stored original88 byte for byte. `format` converts to `glb`, `fbx`, or `obj`; pass `glb` for the LLM-based members,89 which store a Blender file. Ask for `fbx` for a Unity project without a glTF importer, and keep90 the GLB too: at authoring time the FBX embedded only two images, so it may miss PBR maps.9192**Prompt-only build.** Both LLM-based members support `txt23d`. When the user wants no concept to93approve, skip step 1. In step 2, `recommend` with `capability="txt23d"` and the user's words as94`prompt`. Follow steps 3 to 5 with `images: []`; describe the object and name required parts in95`prompt`. A chest's `lid`, `body`, and `padlock` returned as those node names.9697Multi-view models take several images of one subject; read the count and fields off98`model_schema_get`. Side fields follow the subject: on the Tripo multi-view members at authoring99time, `leftImage` was the character's own left (nose toward frame left), `rightImage` the mirror. A100sheet labeled by the viewer's side reverses them, and a swapped pair can fuse front and back into101one shell.102103## Texture and lighting controls104105Several image-to-3D families split texture from geometry. Read each dial in `model_schema_get`106before promising a look. Authoring-time examples from the Tripo members:107108- `texture: false` returns a bare mesh and is a `cost_impact` field: the cheap path when the user109 textures in a DCC. `textureQuality` also moves the price; an empty texture version picker keeps110 the provider's default.111- `delight` strips lighting baked into the reference so the mesh relights in the engine. It112 defaulted to on; turn it off only when painted shading is the art style. The export stays a lit113 material either way: for no relighting at all, the user sets an unlit shader on import.114- `pbr`: the P2 generators defaulted it off (on forces texturing, billed as such); the 3.1115 generators and the texturing member defaulted it on, then ignored the texture parameters. Read its116 description and say which won.117- No Tripo member exposed symmetry: a result unlike the vendor's own app may come from an app-only118 setting.119- Geometry and texture take separate seeds (`seed`, `textureSeed`): hold `seed` and the image, vary120 `textureSeed` for texture variants on one shape.121- `autoSize` scales the output to real-world meters and defaulted to off: set `true` for a122 real-scale import. `dimensions` still reads the unit-normalized mesh (longest side 1.0), since the123 meters sit in the GLB's node transform: not a failure.124125## Inspecting results126127Hosts without MCP Apps get the `app_url` link instead of the viewer. The viewer's capture button128calls `capture_3d_view`, an app-only tool that posts an image `asset_id` of the current view into129the conversation. Reuse it as a generation reference or a similarity `search` image. Never call130`capture_3d_view` yourself: it needs PNG canvas data only the viewer has.131132## Refining meshes1331343D-to-3D utilities retexture, remesh, UV-unwrap, and segment parts: `recommend` with135`capability="3d23d"` plus the operation in the user's own words. Most take the source `asset_id` in136a `kind: "3d"` file field, usually `model` (also `mesh`, `file3d`, `sceneFile`). A 400137`Input model is required` or `Provide a reference image or a 3D model` means the mesh went in under138another name (an image field, or a URL).139140The LLM-based members of step 2 also edit an existing mesh, prop or character:141142- `sceneFile` takes its `asset_id` (any 3D model, returned in the same format), `images` stays143 empty, `prompt` states the change.144- `buildEffort` fits the change, not the object: `standard` for a recolor, `high` or `maximum` to145 add or reshape parts.146- Price follows `buildEffort`, not the mesh, and `recommend` priced no edit at authoring time: a147 `dry_run` on any project mesh quotes it.148- A recolor kept every part name and the face count. A reshape is uncapped (`faceBudget` ignored):149 it took a 4,817-triangle cat to 788,954, so plan a remesh before an engine.150- A replaced material keeps its old name; only added parts get new ones. The edit is generated:151 check what the prompt did not name against the source.152- GPT-6 Astra 3D's description also claimed Mixamo-named rigs, ARKit face shapes, animation loops,153 UV re-unwraps, and modular kits, so it competes in those `recommend` calls.154155Splitting a finished mesh into parts is a contested lane, so it stays a `recommend` pick. At156authoring time vendors offered granularity controls or a bundled PBR retexture, and image-to-parts157members built parts from the picture instead. A convincing mesh is not a game-ready one: part158separation, joints, materials, and animation are each their own pass. Polygon caps differ by159topology: read the slider's `max` instead of promising a count. A named provider feature is a valid160`search` target, but an empty result or one member's enum does not prove a capability absent161platform-wide. `recommend` the unmet need (a hinged prop), inspect the returned schemas, and report162it unverified if none exposes it. Never invent a call from the provider's site.163164## A staged chain toward game-ready165166One image-to-3D run bakes its texture onto UVs the generator chose. A mesh headed for an engine167usually goes further, each output's `asset_id` feeding the next tool's `kind: "3d"` field:1681691. Geometry: a multi-view member when several angles exist, with `texture: false` where offered, and170 `pbr: false` beside it where `pbr` overrides the texture settings.1712. UV unwrap.1723. Retexture the unwrapped mesh from a prompt, an image, or a style reference.1734. Rig last, on the topology that ships.174175Each step after the first is a `3d23d` `recommend` pick and its own billed run: `dry_run` it on the176real input before launching. Along the chain:177178- **Retexture schemas.** Avoid a setting that rebuilds UVs, which discards the unwrap. Check the179 same `pbr` default (the Tripo texturing member defaulted to PBR at authoring time). When a schema180 is silent on UVs, prefer a member with an explicit keep-existing-UVs setting.181- **Input caps.** One unwrapper took meshes up to 30,000 faces at authoring time: set the182 generator's face limit under it (the Tripo multi-view member exposed `faceLimit` and183 `smartLowPoly`), or remesh down first.184- **Checks between steps.** `asset_display` each output. Without a viewer, compare face count, UVs,185 and `dimensions` from `asset_get` step to step, plus the thumbnail asset, which can lag the job by186 seconds. `hasUVs` proves UVs exist, not a clean layout; an unwrapper's "albedo" PNG may be a187 preview, not the atlas.188- **Inherited defects.** Faces rendering dark or inside out are flipped normals, which later steps189 inherit. Scale and pivot can shift between steps, even within one vendor (`autoSize` above):190 confirm both in a DCC before the rig.191- **Sibling formats.** Feed the GLB onward unless the next schema asks otherwise or the mesh is192 quads, which stay in FBX. With no GLB, feed the OBJ or FBX the next schema lists (a `dry_run` may193 skip the format check, so the real run confirms it), else convert with `asset_download` `format`.194195Fixing normals, origin, or naming by hand is DCC work, not an MCP call. The196`scenario-blender-expert` skill drives Blender; it installs locally, so `skill_get` cannot fetch it.197198## Rigging and animation199200Rigging is a separate `3d23d` step on a finished mesh, not a generator flag: `recommend` that201capability plus the rigging need in the user's own words.202203Body plan picks the model. Humanoid riggers take the mesh plus at most a front-facing hint or an204approximate height, and infer a biped skeleton. Non-bipeds go to a model exposing `rigType`:205`quadruped`, `hexapod`, `octopod`, `avian`, `serpentine`, `aquatic`. A static prop has no body plan:206if `recommend` answers `ask_user`, part segmentation is the nearer step.207208A biped headed for a rig starts from a T- or A-pose concept (full body, front view, arms clear of209the torso, plain background): riggers infer joints from the pose. A non-biped keeps step 1's view.210For quads:211212- Pick an image-to-3D member whose schema offers quad topology and set it; the default is triangles.213 At authoring time the Tripo P2 and 3.1 members had a `quad` switch with a lower face cap;214 `scenario-rodin` and `scenario-meshy` name their families' controls.215- Quad output arrives as FBX. glTF stores triangles only, so `format: "glb"` or a GLB-only step216 triangulates it.217- Keep the FBX, rig it with a member whose file field lists FBX, and check each output's face type218 in a DCC before promising quads. A rig returned as GLB is triangulated: deliver the quad FBX219 beside it and say so.220221Three schema details decide whether the output is usable:222223- **Formats.** Riggers accept GLB, often OBJ, FBX, or STL, and expose no output-format field. The224 rig comes back as GLB or FBX, mostly unannounced: expect a DCC pass when the engine needs the225 other.226- **Size ceiling.** A file `max_size` is rare (one humanoid rigger capped at 30 MB): check the227 schema before decimating.228- **Animation versus rig.** Where a rigger exposes `animation`, it retargets a preset clip; its229 `allowed_values` are rig-type prefixed (`quadruped:walk`), so read them. By default only the230 retarget file comes back: set `includeRiggedModel` to keep the plain rig too. The two arrive as231 separate assets with identical `metadata`, in no set order. `asset_get` tells them apart: the232 retarget has `properties.hasAnimations` true and an `animationFrameCount`, the plain rig false and233 null.234235When only motion is wanted, motion-transfer video models animate a still character image with no236skeleton: see `scenario-video`. Video-to-motion models that auto-rig an uploaded mesh are the one237place an `outputFormat` enum picks the engine target.238239## Common mistakes240241- Running `model_run` without `model_schema_get`: 3D parameters differ far more between models than242 image parameters do.243- Hardcoding model IDs: catalogs rotate (a `deprecated:<replacement_id>` tag names the successor).244 Re-discover each session: `recommend` for a need, `search` for a name.245- Promising a named skeleton, an influence count, or bones for wings and extra limbs the chosen246 member does not document: pick the closest body plan, then finish in a DCC.247- Sending a biped to a `rigType` model: the enum has no biped value; humanoids have their own248 riggers.