Generated banner for Omniverse Cad To Simready
Generated logo for Omniverse Cad To Simready

Omniverse Cad To Simready

OpenAIOfficialCreated by OmniverseDistributed by OpenAI

  • Skill
  • Workflow
  • Software development
  • Apache-2.0
  • Medium risk
  • Has scripts

From the author

Coordinate the end-to-end CAD/source-asset to SimReady workflow. Use for broad requests such as CAD to SimReady, source asset to simulation-ready USD, or prop packaging that require conversion, material/physics assignment, SimReady conformance, validation, and optional package creation; deploy or verify Content Agents services first when property assignment is enabled; route single-stage work through nested references.
— the description by Omniverse

From the skill’s files

Taken from the skill’s own files (its description, section headings and stated usage). Not an AutoGineer summary, and not written by the author as one.

Coordinate the end-to-end CAD/source-asset to SimReady workflow. Use for broad requests such as CAD to SimReady, source asset to simulation-ready USD, or prop packaging that require conversion, material/physics assignment, SimReady conformance, validation, and optional package creation; deploy or verify Content Agents services first when property assignment is enabled; route single-stage work through nested references. Use this workflow skill when the user wants an end-to-end pipeline from a source asset to a SimReady asset or package. This skill coordinates existing conversion, authoring, validation, conformance, rendering, and packaging references directly. Do not replace the workflow with a single monolithic runner command.

What it does

  • When to Use
  • Prerequisites
  • Minimum Viable Scope
  • First Action
  • Instructions
  • Output Format

When to use it

  • Use for broad requests such as CAD to SimReady, source asset to simulation-ready USD, or prop packaging that require conversion, material/physics assignment, SimReady conformance, validation, and…
  • Use this workflow skill when the user wants an end-to-end pipeline from a source asset to a SimReady asset or package.
  • This skill coordinates existing conversion, authoring, validation, conformance, rendering, and packaging references directly.
  • This skill is documentation-driven and does not ship scripts/run.py.
  • It should not depend on a repository checkout.
  • When a stage needs deterministic execution, run the portable script from that stage reference's installed directory.

Advantages

The skill’s files do not state this.

Usage examples

The skill’s files do not state this.

Install

  1. Download the skill as a ZIP:

    https://mp-staging.autogineer.com/agent-skills/openai/omniverse-cad-to-simready/download
  2. Unzip it into the cross-client skills folder of your project:

    mkdir -p .agents/skills && unzip omniverse-cad-to-simready.zip -d .agents/skills

Documentation by Omniverse

CAD to SimReady

When to Use

Use this workflow skill when the user wants an end-to-end pipeline from a source asset to a SimReady asset or package. This skill coordinates existing conversion, authoring, validation, conformance, rendering, and packaging references directly. Do not replace the workflow with a single monolithic runner command.

This skill is documentation-driven and does not ship scripts/run.py. It should not depend on a repository checkout. When a stage needs deterministic execution, run the portable script from that stage reference's installed directory. Shell is declared because this workflow invokes installed stage reference scripts directly; it still must not grow a monolithic runner.

Prerequisites

  • Prefer running the preflight reference first for deterministic setup. It installs or verifies local upstream checkouts, writes a cad-to-simready-preflight.json manifest, and exports PHYSICAL_AI_PREFLIGHT_MANIFEST plus PHYSICAL_AI_REQUIRE_PREFLIGHT=1 for downstream references.

  • Python 3.12 and uv (per repo README.md).

  • NVIDIA_API_KEY from https://build.nvidia.com when local Content Agents deployment will run. Already-running endpoints may instead use explicit endpoint variables plus usage tokens such as NGC_API_KEY, NVCF_API_KEY, or CONTENT_AGENTS_*_TOKEN.

  • Docker, NVIDIA Container Toolkit, and an NVIDIA GPU for Content Agents and OVRTX stages.

  • Local upstream checkouts under ${PHYSICAL_AI_SKILL_HUB_UPSTREAM_ROOT:-$HOME/.physical-ai-skill-hub/upstreams} when a downstream stage needs upstream scripts or specs.

Minimum Viable Scope

Conversion-only is a valid workflow request. When the user asks only to convert or smoke-test source asset conversion, set property_assignment_intent=skip, do not deploy Content Agents, run convert-to-usd, then run validate-usd-minimum on the generated USD if conversion succeeds.

Do not imply that uv sync installs every source converter runtime. URDF, MuJoCo/MJCF, and the repo Python dependencies are handled by the project environment, but NVIDIA-backed source conversion requires an installed and validated NVIDIA-Omniverse/usd-convert-cad checkout. If that runtime is missing or does not support the source, preserve the blocked conversion report and its install_hint instead of attempting an unrequested local build or substituting another converter.

First Action

For any broad CAD/source-asset to SimReady request, assume property_assignment_intent=run unless the user explicitly asks for conversion-only, validation-only, or no material/physics assignment.

Before invoking converter, validation, Content Agents, OVRTX, packaging, or FET helper scripts, run the preflight reference or verify an existing PHYSICAL_AI_PREFLIGHT_MANIFEST. Treat preflight as the mandatory dependency bootstrap step, not as workflow routing. If the user explicitly asks not to deploy services or asks for conversion-only/validation-only, use --skip-content-agents. When PHYSICAL_AI_REQUIRE_PREFLIGHT=1 is set and a required component is not ready in the manifest, downstream references must block with the preflight guardrail instead of rediscovering upstreams or services directly.

When property_assignment_intent=run, the first operational action after confirming the source path and resolving intent is to verify or deploy Content Agents services. Do this before asset-context inspection, converter dependency checks, conversion, validation, conformance, rendering, packaging, or upstream source builds.

Use healthy existing endpoints when available. If OVRTX, Material, or Physics endpoints are missing or unhealthy, run deploy-content-agents first and do not continue until the shared standalone OVRTX renderer plus independent Material and Physics service containers are healthy and exported through CONTENT_AGENTS_*_BASE_URL. Deploy the Texture Agent too when texture generation is requested.

If required deployment authentication is missing, ask the user for NVIDIA_API_KEY and wait. If a provided endpoint requires usage auth, ask for the appropriate usage token instead. If deployment cannot produce healthy services, report Content Agents readiness as blocked instead of proceeding to conversion.

Instructions

  1. Confirm the source asset path exists, resolve output_root, and classify the request as end-to-end, conversion-only, validation-only, or packaging.

  2. Resolve property_assignment_intent before running any asset inspection, converter probe, conversion, validation, conformance, rendering, or packaging step.

  3. Run preflight for the selected workflow targets, unless a ready PHYSICAL_AI_PREFLIGHT_MANIFEST is already configured. Source the generated env file before running downstream scripts. Treat preflight as dependency setup only: it may use a provided --source-asset, --source-format, or --conversion-tools value to scope dependency checks, but convert-to-usd and the upstream converter references still decide actual conversion support.

  4. Verify or deploy Content Agents services first when property_assignment_intent=run; block on missing authentication or unhealthy services instead of continuing.

  5. Read references/workflow.md and references/commands.md, then run only the stage references needed for the current request.

  6. Run identify-asset-context on the original source asset when web search is available or property assignment will run.

  7. Route the source through convert-to-usd, or skip conversion for existing USD input and treat the source path as the current USD path.

  8. Run validate-usd-minimum before expensive downstream work. Treat this as a viability gate only: record unit/profile issues such as metersPerUnit != 1.0, but do not run simready-conform-profile, FET001, or any other FET repair before Content Agents assignment when property assignment will run.

  9. Run Content Agents material, physics, and optional texture assignment on the converted/minimum-valid USD when requested or required.

  10. Run simready-conform-profile on the latest simulation USD path after property assignment and preserve every selected FET repair report.

  11. Run validation gates in order: omni-asset-validate, omni-asset-validate-geometry, omni-asset-validate-physics, and simready-validate.

  12. Rerun simready-conform-profile when simready-validate reports a repairable requirement, then rerun profile validation on the newest authored USD.

  13. Run ovrtx-render-service when preview, thumbnail, or inspection images are requested. When package outputs are requested, run assemble-package-source next to create the clean deliverable/ package source from the final USD and thumbnail, then run nv-core-package-sample and nv-core-package-sample-validation on that deliverable folder only.

  14. Emit the consolidated workflow report with the final USD path, all stage reports, validation findings, rerun reasons, and next work.

Use the simready-conform-profile reference only after property assignment when property_assignment_intent=run. It routes feature repair to upstream SimReady Foundation FET skills such as simready-foundation-conform-fet-000-core, simready-foundation-conform-fet-001-minimal, simready-foundation-conform-fet-004-simulate-multi-body-physics, and simready-foundation-conform-fet-005-simulate-grasp-physics from branch main.

If simready-validate reports a repairable requirement after the first conformance pass, feed the structured requirement IDs back into the simready-conform-profile reference before writing the final result. In particular, GSP.001 is owned by upstream simready-foundation-conform-fet-005-simulate-grasp-physics; run that skill when a vision-capable agent can inspect visual evidence or explicit grasp points were provided, otherwise record the FET005 step as blocked by missing vision/points instead of treating it as an optional preview task. For RB.MB.001, route the failure to upstream simready-foundation-conform-fet-004-simulate-multi-body-physics. Do not assume multiple visual prims are multiple rigid bodies; inspect UsdPhysics.RigidBodyAPI applications. When the Physics Agent report shows composed topology optimization or the USD has existing component colliders/part roots and the profile validator reports FET004/RB.MB.001, FET004 should promote those existing components into rigid bodies without creating geometry. Do not mark the gate not applicable until after confirming there are fewer than two reusable body candidates.

Output Format

Emit a consolidated workflow report in Markdown, and include JSON when the workflow writes structured artifacts. The report must include:

  • Overall status: passed, blocked, failed, or needs_rerun.

  • Request summary: source asset path, detected source format, output root, selected SimReady profile/version, and property assignment intent.

  • Ordered stage results: stage reference, input artifact, output USD or USDZ path, report path, status, blocker reason, and rerun reason when applicable.

  • Content Agents readiness and property assignment results with service URLs, tokens, and credentials redacted.

  • Conformance and validation findings grouped by gate, requirement ID, selected FET repair reference, repair-loop attempt, and final disposition.

  • Final artifacts: final reported USD path, render preview path when requested, package root and package validation report when packaging ran, Markdown report path, JSON report path when present, and recommended next work.

Detailed References

Read only the references needed for the current request:

  • references/preflight/README.md: deterministic local setup, manifest/env contract, Linux and Windows wrappers, Content Agents deployment opt-out, and guardrail behavior.

  • references/workflow.md: inputs, source routing, detailed workflow, validation policy, output report fields, approval points, and next steps.

  • references/commands.md: concrete portable script command patterns for each stage.

  • references/assemble-package-source/README.md: two-zone package source assembly, canonical root USD naming, thumbnail placement, and self-contained deliverable checks.

Publishing Layout Notes

Use skills/omniverse-cad-to-simready/ as the source of truth for this product repo's skill. The .agents/skills symlink is a compatibility alias for local agentskills.io-style discovery, and .codex/skills and .claude/skills are agent-specific compatibility aliases.

Frontmatter keeps version and tools at top level for agentskills.io runtime compatibility. NVCARPS discoverability fields live under metadata.

The nested references/ tree is intentional. It keeps one public catalog skill while retaining script-bearing atomic stage references, upstream handoff notes, and router documentation under the workflow. Do not flatten those references or promote nested README references to sibling SKILL.md files unless the repo's publishing model changes.

Limitations

  • This workflow coordinates existing conversion, property assignment, conformance, validation, rendering, and packaging skills; it does not replace them with a single monolithic runner command.

  • Stop at the first failing deployment, conversion, property-assignment, or conformance authoring gate unless the user explicitly asks for best-effort continuation.

  • Upstream simready-foundation-conform-fet-005-simulate-grasp-physics needs visual review or explicit grasp points before it can author a meaningful grasp vector.

Troubleshooting

SymptomCauseFix
Downstream reference reports that cad-to-simready preflight has not prepared a componentPHYSICAL_AI_REQUIRE_PREFLIGHT=1 is set, but the manifest is missing or the required runtime/service is not readyRun preflight/scripts/preflight.py, source the generated env file, or explicitly disable service deployment with --skip-content-agents only when Content Agents are out of scope.
Workflow stops on GSP.001 and reports the failure as unclassifiedVisual evidence or explicit grasp points were not provided to FET005Run upstream simready-foundation-conform-fet-005-simulate-grasp-physics only after a vision-capable agent has reviewed the asset, or pass explicit grasp points. Otherwise report the FET005 step as blocked, not failed.
Validation fails after a meaningful USD artifact already existsWorkflow stopped at the first validation findingContinue remaining diagnostic gates and mark the result needs_rerun. Do not stop at validation findings once a USD artifact has been produced.
Property-assignment stage fails with a missing service endpointContent Agents service was not deployed before conversionRun deploy-content-agents first. Do not start asset inspection, conversion, validation, conformance, rendering, or packaging before Content Agents readiness when property assignment will run.
Material Agent reports that rendering produced 0 images after unit or profile repairA FET repair, commonly FET001 unit normalization, was applied before Material Agent and changed the USD layering/scene state consumed by the serviceRerun assignment from the converted/minimum-valid USD: Material Agent first, then Physics Agent, then run simready-conform-profile and FET repairs on the latest service-authored USD.
Material or Physics Agent local optimized path reports Permission denied: '/app/.build-resources/scene_optimizer_core/python'Local Docker Scene Optimizer bundle permissions prevent the non-root service user from reading the packaged SO runtimeRepair the relevant local container with docker exec --user root content-material-agent-service chmod -R a+rX /app/.build-resources/scene_optimizer_core or docker exec --user root content-physics-agent-service chmod -R a+rX /app/.build-resources/scene_optimizer_core, then rerun the same optimized agent command. Do not treat the no-optimizer fallback as the root cause for instanced/prototype assets.
RB.MB.001 fails even though the asset has many primsThe profile counts UsdPhysics.RigidBodyAPI prims, not visual or collider prims; Physics Agent may author one root rigid bodyRoute to upstream simready-foundation-conform-fet-004-simulate-multi-body-physics. First ensure Physics Agent used composed-topology optimization when applicable, then promote existing component colliders/part roots when the active profile reports FET004/RB.MB.001 and no geometry must be invented.

Hard Rules

  • Prefer the preflight manifest for local upstream roots, converter executables, SimReady validation runtime, OVRTX endpoint, and Content Agents service URLs. When PHYSICAL_AI_REQUIRE_PREFLIGHT=1 is set, do not bypass the manifest with direct upstream discovery.

  • Do not run asset inspection, converter probes, local upstream builds, conversion, validation, conformance, rendering, or packaging before Content Agents readiness when property assignment will run.

  • Use stage-specific installed reference scripts directly. Do not add or call a single omniverse-cad-to-simready runner command.

  • For source conversion, delegate to the convert-to-usd reference; do not substitute another converter for CAD or mesh formats.

  • For property assignment, use Content Agents references as separate atomic steps: material first, then physics, then texture only when requested.

  • When property assignment will run, do not run simready-conform-profile or any FET helper before Content Agents. Validate minimum USD first, then run Content Agents on that converted/minimum-valid USD, then apply FET repairs to the latest service-authored USD.

  • When property assignment will run, do not run simready-validate or any SimReady profile validation before Content Agents. The only validation gate allowed before service calls is validate-usd-minimum, which is a basic USD viability check.

  • Stop at the first failing deployment, conversion, property-assignment, or conformance authoring gate unless the user explicitly asks for best-effort continuation.

  • Do not stop at validation findings after a meaningful USD artifact exists. Continue remaining diagnostic gates and mark the result needs_rerun.

  • Do not leave a GSP.001 profile failure as an unclassified final finding. Route it to upstream simready-foundation-conform-fet-005-simulate-grasp-physics; if the current agent cannot inspect renders or no explicit grasp points are available, report a blocked FET005 repair with the visual evidence path or missing input reason.

  • Preserve every stage report and pass the concrete output USD path from each report into the next stage.

Files

references/convert-to-usd/scripts/run.py

#!/usr/bin/env python3
# SPDX-FileCopyrightText: Copyright (c) 2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
# SPDX-License-Identifier: Apache-2.0

"""Route a source asset by asking upstream converter references for support.

Usage:
    python3 scripts/run.py <source_asset> <output_directory> [--report PATH]
    python3 scripts/run.py <source_asset> <output_directory> [--markdown-report PATH]

Arguments:
    source_asset            Path to the input file or directory to convert.
    output_directory        Directory to write the generated USD artifact and reports.
    --report PATH           Optional path to write the normalized conversion report (JSON).
    --markdown-report PATH  Optional path to write the normalized conversion report (Markdown).

Exit codes:
    0 - conversion succeeded or source is already USD
    1 - expected failure (unsupported source format or missing required converter)
    2 - unexpected error (crash or malformed input)
"""
from __future__ import annotations

import argparse
from dataclasses import dataclass, field, replace
import json
from pathlib import Path
import subprocess
import sys
import tempfile
from typing import Any
import zipfile


SKILL = "convert-to-usd"
NEXT_STEP = "validate-usd-minimum"
PROBE_TIMEOUT_SECONDS = 30
CONVERSION_TIMEOUT_SECONDS = 1900
REFERENCE_ORDER = (
    "urdf-usd-converter",
    "mujoco-usd-converter",
    "usd-convert-gsplat",
    "usd-convert-cad",
)


@dataclass(frozen=True)
class ProbeResult:
    converter_skill: str
    converter_tool: str
    source_format: str
    supported: bool
    warnings: list[str] = field(default_factory=list)
    errors: list[str] = field(default_factory=list)
    install_hint: str = ""


@dataclass(frozen=True)
class SourceSelection:
    source_asset: Path
    warnings: list[str] = field(default_factory=list)
    selected_probe: ProbeResult | None = None
    probes: list[ProbeResult] = field(default_factory=list)


@dataclass(frozen=True)
class ConversionReport:
    source_asset_path: str
    source_format: str
    converter_skill: str
    converter_tool: str
    converter_command: list[str]
    output_directory: str
    output_usd_path: str
    generated_files: list[str]
    converter_reference: str = ""
    sidecar_inputs: list[str] = field(default_factory=list)
    warnings: list[str] = field(default_factory=list)
    errors: list[str] = field(default_factory=list)
    diagnostics: list[dict[str, Any]] = field(default_factory=list)
    install_hint: str = ""
    next_step: str = NEXT_STEP

    @property
    def passed(self) -> bool:
        return not self.errors

    def to_dict(self) -> dict[str, Any]:
        payload = {
            "source_asset_path": self.source_asset_path,
            "source_format": self.source_format,
            "converter_skill": self.converter_skill,
            "converter_tool": self.converter_tool,
            "converter_command": self.converter_command,
            "output_directory": self.output_directory,
            "output_usd_path": self.output_usd_path,
            "generated_files": self.generated_files,
            "sidecar_inputs": self.sidecar_inputs,
            "warnings": self.warnings,
            "errors": self.errors,
            "next_step": self.next_step,
        }
        if self.converter_reference:
            payload["converter_reference"] = self.converter_reference
        if self.diagnostics:
            payload["diagnostics"] = self.diagnostics
        if self.install_hint:
            payload["install_hint"] = self.install_hint
        return payload

    def to_json(self) -> str:
        return json.dumps(self.to_dict(), indent=2, sort_keys=True) + "\n"

    def to_markdown(self) -> str:
        command = " ".join(self.converter_command)
        lines = [
            "# Conversion Report",
            "",
            f"- Source asset: `{self.source_asset_path}`",
            f"- Source format: `{self.source_format}`",
            f"- Converter skill: `{self.converter_skill}`",
            f"- Converter tool: `{self.converter_tool}`",
            f"- Converter command: `{command}`",
            f"- Output directory: `{self.output_directory}`",
            f"- Output USD: `{self.output_usd_path}`",
            f"- Next step: `{self.next_step}`",
            "",
            "## Generated Files",
            "",
        ]
        lines.extend(f"- `{path}`" for path in self.generated_files)
        if not self.generated_files:
            lines.append("- None")
        lines.extend(["", "## Warnings", ""])
        lines.extend(f"- {warning}" for warning in self.warnings)
        if not self.warnings:
            lines.append("- None")
        lines.extend(["", "## Errors", ""])
        lines.extend(f"- {error}" for error in self.errors)
        if not self.errors:
            lines.append("- None")
        if self.install_hint:
            lines.extend(["", "## Install Hint", "", self.install_hint])
        lines.append("")
        return "\n".join(lines)


def reference_root() -> Path:
    return Path(__file__).resolve().parents[1] / "references"


def reference_run_script(converter_skill: str) -> Path:
    return reference_root() / converter_skill / "scripts" / "run.py"


def is_existing_usd(source_asset: Path) -> bool:
    if not source_asset.is_file():
        return False
    try:
        with source_asset.open("rb") as file:
            header = file.read(16)
    except OSError:
        return False
    if header.startswith(b"#usda") or header.startswith(b"PXR-USDC"):
        return True
    if zipfile.is_zipfile(source_asset):
        try:
            with zipfile.ZipFile(source_asset) as archive:
                for name in archive.namelist():
                    with archive.open(name) as member:
                        member_header = member.read(16)
                    if member_header.startswith(b"#usda") or member_header.startswith(b"PXR-USDC"):
                        return True
                return False
        except zipfile.BadZipFile:
            return False
    return False


def already_usd_report(source_asset: Path, output_directory: Path) -> ConversionReport:
    return ConversionReport(
        source_asset_path=str(source_asset),
        source_format="usd",
        converter_skill=SKILL,
        converter_tool="none",
        converter_command=[],
        output_directory=str(output_directory),
        output_usd_path=str(source_asset),
        generated_files=[],
        warnings=["Source asset is already USD; conversion skipped"],
    )


def missing_source_report(source_asset: Path, output_directory: Path) -> ConversionReport:
    return ConversionReport(
        source_asset_path=str(source_asset),
        source_format="unknown",
        converter_skill=SKILL,
        converter_tool="none",
        converter_command=[],
        output_directory=str(output_directory),
        output_usd_path="",
        generated_files=[],
        errors=["Source asset does not exist"],
    )


def run_probe(source_asset: Path, converter_skill: str) -> ProbeResult:
    script = reference_run_script(converter_skill)
    if not script.exists():
        return ProbeResult(
            converter_skill=converter_skill,
            converter_tool="none",
            source_format="unknown",
            supported=False,
            errors=[f"converter reference script is missing: {script}"],
        )

    command = [sys.executable, str(script), str(source_asset), "--probe"]
    try:
        completed = subprocess.run(
            command,
            capture_output=True,
            text=True,
            timeout=PROBE_TIMEOUT_SECONDS,
            check=False,
        )
    except subprocess.TimeoutExpired:
        return ProbeResult(
            converter_skill=converter_skill,
            converter_tool=converter_skill,
            source_format="unknown",
            supported=False,
            errors=[f"converter probe timed out after {PROBE_TIMEOUT_SECONDS}s: {converter_skill}"],
        )

    try:
        payload = json.loads(completed.stdout)
    except json.JSONDecodeError:
        detail = completed.stderr.strip() or completed.stdout.strip() or f"exit {completed.returncode}"
        return ProbeResult(
            converter_skill=converter_skill,
            converter_tool=converter_skill,
            source_format="unknown",
            supported=False,
            errors=[f"converter probe did not return JSON for {converter_skill}: {detail}"],
        )

    return ProbeResult(
        converter_skill=str(payload.get("converter_skill") or converter_skill),
        converter_tool=str(payload.get("converter_tool") or converter_skill),
        source_format=str(payload.get("source_format") or "unknown"),
        supported=bool(payload.get("supported")),
        warnings=[str(warning) for warning in payload.get("warnings", [])],
        errors=[str(error) for error in payload.get("errors", [])],
        install_hint=str(payload.get("install_hint") or ""),
    )


def probe_warnings(probes: list[ProbeResult]) -> list[str]:
    warnings: list[str] = []
    for probe in probes:
        status = "supported" if probe.supported else "not supported"
        warnings.append(f"Probe {probe.converter_skill}: {status} ({probe.source_format})")
        warnings.extend(f"{probe.converter_skill}: {warning}" for warning in probe.warnings)
        warnings.extend(f"{probe.converter_skill}: {error}" for error in probe.errors)
    return warnings


def select_converter(source_asset: Path) -> tuple[str | None, ProbeResult | None, list[ProbeResult]]:
    probes = [run_probe(source_asset, converter_skill) for converter_skill in REFERENCE_ORDER]
    supported = [probe for probe in probes if probe.supported]
    if not supported:
        return None, None, probes
    selected = supported[0]
    return selected.converter_skill, selected, probes


def source_relative_label(path: Path, root: Path) -> str:
    try:
        return path.relative_to(root).as_posix()
    except ValueError:
        return str(path)


def supported_source_label(selection: SourceSelection, root: Path) -> str:
    label = source_relative_label(selection.source_asset, root)
    if selection.selected_probe is None:
        return f"`{label}` (existing USD)"
    return f"`{label}` ({selection.selected_probe.converter_skill}, {selection.selected_probe.source_format})"


def report_with_warnings(report: ConversionReport, warnings: list[str]) -> ConversionReport:
    if not warnings:
        return report
    return replace(report, warnings=[*warnings, *report.warnings])


def directory_inspection_report(source_directory: Path, output_directory: Path, error: str) -> ConversionReport:
    return ConversionReport(
        source_asset_path=str(source_directory),
        source_format="unknown",
        converter_skill=SKILL,
        converter_tool="none",
        converter_command=[],
        output_directory=str(output_directory),
        output_usd_path="",
        generated_files=[],
        errors=[error],
    )


def unsupported_directory_report(
    source_directory: Path,
    output_directory: Path,
    inspected_count: int,
    probes: list[ProbeResult],
) -> ConversionReport:
    install_hint = next((probe.install_hint for probe in probes if probe.install_hint), "")
    if inspected_count == 0:
        detail = "no files"
        error = "directory source is empty; expected exactly one supported source file"
    else:
        detail = f"{inspected_count} file(s)"
        error = (
            "directory source does not contain a supported source file among "
            f"{inspected_count} inspected file(s)"
        )
    return ConversionReport(
        source_asset_path=str(source_directory),
        source_format="unknown",
        converter_skill=SKILL,
        converter_tool="none",
        converter_command=[],
        output_directory=str(output_directory),
        output_usd_path="",
        generated_files=[],
        warnings=[f"Inspected directory source `{source_directory}` and found {detail}."],
        errors=[error],
        install_hint=install_hint,
    )


def ambiguous_directory_report(
    source_directory: Path,
    output_directory: Path,
    selections: list[SourceSelection],
) -> ConversionReport:
    candidates = ", ".join(supported_source_label(selection, source_directory) for selection in selections)
    return ConversionReport(
        source_asset_path=str(source_directory),
        source_format="unknown",
        converter_skill=SKILL,
        converter_tool="none",
        converter_command=[],
        output_directory=str(output_directory),
        output_usd_path="",
        generated_files=[],
        errors=[
            "directory source is ambiguous because multiple supported source files were found: "
            f"{candidates}. Pass one source file explicitly."
        ],
    )


def select_directory_source(
    source_directory: Path,
    output_directory: Path,
) -> tuple[SourceSelection | None, ConversionReport | None]:
    try:
        files = sorted(
            (path for path in source_directory.rglob("*") if path.is_file()),
            key=lambda path: source_relative_label(path, source_directory),
        )
    except OSError as exc:
        return None, directory_inspection_report(
            source_directory,
            output_directory,
            f"could not inspect directory source: {exc}",
        )

    selections: list[SourceSelection] = []
    probes: list[ProbeResult] = []
    for path in files:
        if is_existing_usd(path):
            selections.append(SourceSelection(source_asset=path))
            continue
        _, selected_probe, source_probes = select_converter(path)
        probes.extend(source_probes)
        if selected_probe is not None:
            selections.append(
                SourceSelection(source_asset=path, selected_probe=selected_probe, probes=source_probes)
            )

    if not selections:
        return None, unsupported_directory_report(source_directory, output_directory, len(files), probes)
    if len(selections) > 1:
        return None, ambiguous_directory_report(source_directory, output_directory, selections)

    selection = selections[0]
    label = source_relative_label(selection.source_asset, source_directory)
    warning = f"Directory source contained exactly one supported source file; selected `{label}` for conversion."
    return replace(selection, warnings=[warning]), None


def run_converter(source_asset: Path, output_directory: Path, converter_skill: str) -> ConversionReport:
    script = reference_run_script(converter_skill)
    with tempfile.TemporaryDirectory(prefix="convert-to-usd-") as temp_dir:
        report_path = Path(temp_dir) / "conversion.json"
        command = [
            sys.executable,
            str(script),
            str(source_asset),
            str(output_directory),
            "--report",
            str(report_path),
        ]
        try:
            completed = subprocess.run(
                command,
                capture_output=True,
                text=True,
                timeout=CONVERSION_TIMEOUT_SECONDS,
                check=False,
            )
        except subprocess.TimeoutExpired:
            return ConversionReport(
                source_asset_path=str(source_asset),
                source_format="unknown",
                converter_skill=converter_skill,
                converter_reference=converter_skill,
                converter_tool=converter_skill,
                converter_command=command,
                output_directory=str(output_directory),
                output_usd_path="",
                generated_files=[],
                errors=[f"converter timed out after {CONVERSION_TIMEOUT_SECONDS}s: {converter_skill}"],
            )

        if report_path.exists():
            try:
                payload = json.loads(report_path.read_text(encoding="utf-8"))
            except json.JSONDecodeError as exc:
                return malformed_converter_report(source_asset, output_directory, converter_skill, command, f"invalid JSON report: {exc}")
        else:
            detail = completed.stderr.strip() or completed.stdout.strip() or f"exit {completed.returncode}"
            return malformed_converter_report(source_asset, output_directory, converter_skill, command, f"missing converter report: {detail}")

    return report_from_payload(payload, converter_skill)


def malformed_converter_report(
    source_asset: Path,
    output_directory: Path,
    converter_skill: str,
    command: list[str],
    error: str,
) -> ConversionReport:
    return ConversionReport(
        source_asset_path=str(source_asset),
        source_format="unknown",
        converter_skill=converter_skill,
        converter_reference=converter_skill,
        converter_tool=converter_skill,
        converter_command=command,
        output_directory=str(output_directory),
        output_usd_path="",
        generated_files=[],
        errors=[error],
    )


def report_from_payload(payload: dict[str, Any], converter_skill: str) -> ConversionReport:
    return ConversionReport(
        source_asset_path=str(payload.get("source_asset_path") or ""),
        source_format=str(payload.get("source_format") or "unknown"),
        converter_skill=str(payload.get("converter_skill") or converter_skill),
        converter_reference=str(payload.get("converter_reference") or payload.get("converter_skill") or converter_skill),
        converter_tool=str(payload.get("converter_tool") or converter_skill),
        converter_command=[str(part) for part in payload.get("converter_command", [])],
        output_directory=str(payload.get("output_directory") or ""),
        output_usd_path=str(payload.get("output_usd_path") or ""),
        generated_files=[str(path) for path in payload.get("generated_files", [])],
        sidecar_inputs=[str(path) for path in payload.get("sidecar_inputs", [])],
        warnings=[str(warning) for warning in payload.get("warnings", [])],
        errors=[str(error) for error in payload.get("errors", [])],
        diagnostics=[diagnostic for diagnostic in payload.get("diagnostics", []) if isinstance(diagnostic, dict)],
        install_hint=str(payload.get("install_hint") or ""),
        next_step=str(payload.get("next_step") or NEXT_STEP),
    )


def append_selection_warnings(report: ConversionReport, selected: ProbeResult, probes: list[ProbeResult]) -> ConversionReport:
    supported = [probe.converter_skill for probe in probes if probe.supported]
    warnings = list(report.warnings)
    warnings.append(f"Router selected `{selected.converter_skill}` from upstream converter capability probes.")
    if len(supported) > 1:
        warnings.append(
            "Multiple converter references reported support; selected by converter-reference priority: "
            + ", ".join(supported)
        )
    return ConversionReport(
        source_asset_path=report.source_asset_path,
        source_format=report.source_format,
        converter_skill=report.converter_skill,
        converter_reference=report.converter_reference,
        converter_tool=report.converter_tool,
        converter_command=report.converter_command,
        output_directory=report.output_directory,
        output_usd_path=report.output_usd_path,
        generated_files=report.generated_files,
        sidecar_inputs=report.sidecar_inputs,
        warnings=warnings,
        errors=report.errors,
        diagnostics=report.diagnostics,
        install_hint=report.install_hint,
        next_step=report.next_step,
    )


def unsupported_report(source_asset: Path, output_directory: Path, probes: list[ProbeResult]) -> ConversionReport:
    install_hint = next((probe.install_hint for probe in probes if probe.install_hint), "")
    return ConversionReport(
        source_asset_path=str(source_asset),
        source_format="unknown",
        converter_skill=SKILL,
        converter_tool="none",
        converter_command=[],
        output_directory=str(output_directory),
        output_usd_path="",
        generated_files=[],
        warnings=probe_warnings(probes),
        errors=["no converter reference reported support for this source asset"],
        install_hint=install_hint,
    )


def convert_to_usd(source_asset: Path, output_directory: Path) -> ConversionReport:
    source_asset = source_asset.resolve()
    output_directory = output_directory.resolve()
    directory_warnings: list[str] = []
    selected_probe: ProbeResult | None = None
    probes: list[ProbeResult] = []

    if not source_asset.exists():
        return missing_source_report(source_asset, output_directory)
    if source_asset.is_dir():
        selection, report = select_directory_source(source_asset, output_directory)
        if report is not None:
            return report
        if selection is None:
            return unsupported_directory_report(source_asset, output_directory, 0, [])
        source_asset = selection.source_asset
        directory_warnings = selection.warnings
        selected_probe = selection.selected_probe
        probes = selection.probes
    if is_existing_usd(source_asset):
        return report_with_warnings(already_usd_report(source_asset, output_directory), directory_warnings)

    if selected_probe is None:
        converter_skill, selected_probe, probes = select_converter(source_asset)
    else:
        converter_skill = selected_probe.converter_skill
    if converter_skill is None or selected_probe is None:
        return unsupported_report(source_asset, output_directory, probes)

    report = run_converter(source_asset, output_directory, converter_skill)
    return report_with_warnings(append_selection_warnings(report, selected_probe, probes), directory_warnings)


def emit_report(
    report: ConversionReport,
    *,
    report_path: Path | None = None,
    markdown_report_path: Path | None = None,
) -> None:
    report_json = report.to_json()
    if report_path is not None:
        report_path.parent.mkdir(parents=True, exist_ok=True)
        report_path.write_text(report_json, encoding="utf-8")
    if markdown_report_path is not None:
        markdown_report_path.parent.mkdir(parents=True, exist_ok=True)
        markdown_report_path.write_text(report.to_markdown(), encoding="utf-8")
    print(report_json, end="")


def main(argv: list[str] | None = None) -> int:
    parser = argparse.ArgumentParser(description="Route a source asset by querying upstream converter references.")
    parser.add_argument("source_asset", type=Path)
    parser.add_argument("output_directory", type=Path)
    parser.add_argument("--report", type=Path, help="Write a JSON report to this path.")
    parser.add_argument("--markdown-report", type=Path, help="Write a Markdown report to this path.")
    args = parser.parse_args(argv)

    report = convert_to_usd(args.source_asset, args.output_directory)
    emit_report(report, report_path=args.report, markdown_report_path=args.markdown_report)
    return 0 if report.passed else 1


if __name__ == "__main__":
    raise SystemExit(main())

Requirements and compatibility

Written for: OpenAI Codex

Works with: Claude Code, Claude apps, OpenAI Codex, GitHub Copilot, Cursor, Gemini CLI, Kiro, OpenCode, Goose, Amp, Roo Code, JetBrains Junie, Factory Droid, Devin (ex-Windsurf), Mistral Vibe, OpenHands, Letta

“Works with” means the tool reads the open Agent Skills format; a skill written for one tool may use features others lack.

Needs: Needs an API key, Needs Python

The author says: Orchestrator skill. Managed Content Agents deployment requires NVIDIA_API_KEY (build.nvidia.com), Docker + NVIDIA Container Toolkit + GPU, Python 3.12, and an upstream checkout of nvidia-omniverse/content-agents on branch main. Reused/provided endpoints may instead use explicit endpoint and usage-token environment variables. Linux/macOS only.

Security

What our scanner noticed· Medium risk

  • Large base64-encoded blobs
  • Instructions that use the network
  • Downloads piped into a shell
  • Scripts that your agent may run

This skill includes scripts that your agent may run. Read them before you use it.

A static scan is not a guarantee. Skills are third-party content: review a skill before you let an agent use it.

Attribution and licence

Created by
Omniverse
Publisher
OpenAI (Official)
Distributed by
OpenAI
Source
openai/plugins · plugins/nvidia/skills/omniverse-cad-to-simready · commit d420d13
Licence
Apache License 2.0 (redistribution allowed)
  • LICENSE (written by AutoGineer)
  • ATTRIBUTION.md (written by AutoGineer)
Synced

AutoGineer is not affiliated with or endorsed by Omniverse or OpenAI.

Download ZIP

The ZIP holds the skill’s files, its licence files and an ATTRIBUTION.md. Download omniverse-cad-to-simready.zip