CLI & Python SDK

View Markdown llms-full.txt

rigyd is the official command-line tool and Python SDK for the Rigyd API. Version 1.1 adds the same SimReady Asset Composer workflow used by the public API: start from intent and optional reference images, or convert an existing 3D model. The core package has zero dependencies.

pip install https://github.com/rigyd/rigyd-cli/releases/download/v1.1.0/rigyd-1.1.0-py3-none-any.whl

Version 1.1.0 is available from the GitHub release. The PyPI package remains on 1.0.1 until the new release is published there.

CLI

Authenticate. Stores your rgyd_live_... key at ~/.config/rigyd/config.json (mode 600).
Terminal window
```
rigyd login
```
Key resolution order: --api-key flag → RIGYD_API_KEY env → config file.
Compose from an idea or reference image.

rigyd compose creates the composition, polls it through prototype generation and SimReady conversion, then downloads the result. Unlike the REST API, this one-command path enables automatic submission by default.
Terminal window
```
rigyd compose "compact desktop stapler" \
  --task "press the upper arm to staple paper" \
  --asset-class articulated \
  --export all
```
Add one --image reference.png, or four ordered --image flags for front, right, back, and left views. Use --asset-class auto when Rigyd should infer whether the asset is rigid or articulated.
Pause for review when needed.
Terminal window
```
rigyd compose "countertop microwave" \
  --task "open the door and place a container inside" \
  --review

rigyd compositions get <composition_id>
rigyd compositions submit <composition_id> --export all
```
--review maps to auto_submit=false. The command returns the composition ID when the prototype reaches ready; submitting it later creates and downloads the linked conversion job.

Or convert an existing 3D asset.

# 3D file → SimReady, optionally retopologized
rigyd convert chair.glb --tris 50000 --export all

Inspect, re-download, or simulate.

rigyd compositions list
rigyd jobs list
rigyd jobs get <job_id>
rigyd download <job_id> --export mujoco   # re-download any job, 0 credits
rigyd simulate <job_id> --scene drop      # physics demo video, 0 credits
rigyd whoami                              # user + credit balance

--export takes a format (usd, mjcf, all) or a simulator alias (isaac → USD, mujoco → MJCF). Default: usd.
Progress goes to stderr, the result path to stdout, so it composes with other tools: blender $(rigyd convert scan.obj --export usd). Add --json for a machine-readable manifest (handy for agents and CI).

Input	Command	Cost
Idea + optional 1 or 4 images	`rigyd compose PROMPT --task TASK`	Prototype price from `rigyd pricing`
3D file (`.glb/.gltf/.fbx/.obj/.stl/.ply/.usd*`)	`rigyd convert FILE`	Free

See Supported formats for the full input list and Pricing for credit costs.

Python SDK

import rigyd
rigyd.configure()                                  # key from login / RIGYD_API_KEY

job = rigyd.convert(file="chair.glb")
job.wait(on_progress=lambda j: print(j.status, j.stage, j.progress))
usd_path = job.download(fmt="usd")                 # or "mjcf" / "all"
print(rigyd.account())                             # user + credit balance

Compose and automatically continue into SimReady conversion:

composition = rigyd.compose(
    prompt="compact desktop stapler",
    robot_task="press the upper arm to staple paper",
    asset_class="articulated",                    # auto | rigid | articulated
    auto_submit=True,
    physical_context={
        "dimensions_m": {"length": 0.15, "width": 0.04, "height": 0.08},
        "materials": ["ABS plastic", "steel"],
        "fixed_base": False,
    },
)
composition.wait()
job = composition.conversion_job()
usd_path = job.download(fmt="usd")

The SDK follows the API default: auto_submit=False. Call composition.submit() after review, or set it to True for an unattended run.

composition.wait() polls the composition workflow until ready, completed, failed, or cancelled. job.wait() polls the linked conversion lifecycle and raises on failure.

MuJoCo extra

Install the optional dependency to load a result straight into MuJoCo:

pip install "rigyd[mujoco]"

model = rigyd.load_model(file="chair.glb")  # -> mujoco.MjModel, ready to mj_step

Loading into a live simulator

NVIDIA Isaac Sim — use the Isaac Sim extension to convert an existing 3D asset and load it onto the stage without leaving the app. Use this CLI for text/image composition or offline and scripted pipelines.
MuJoCo — rigyd.load_model(...) above, or mujoco.MjModel.from_xml_path(<path from rigyd download --export mujoco>).

Links

Source & full reference github.com/rigyd/rigyd-cli

Version 1.1.0 release Composer-enabled wheel and source archive