Generated banner for Omniverse Usd Performance Tuning
Generated logo for Omniverse Usd Performance Tuning

Omniverse Usd Performance Tuning

NVIDIAOfficialCreated by NVIDIA Omniverse

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

From the author

Diagnose and optimize USD stages for slow loading, high memory, low FPS or validation; preserve asset boundaries and reuse shared definitions.
— the description by NVIDIA Omniverse

AutoGineer summary

Written by AutoGineer from the skill’s own files, not by the author

This skill provides a structured workflow for diagnosing and optimizing USD stages. It focuses on improving performance metrics like load times, memory footprint, and frame rates by identifying and optimizing unique asset definitions, reusing shared definitions, and pruning unnecessary data. The process emphasizes preserving original asset boundaries, animation, and interface requirements, ensuring that downstream applications remain functional after optimization.

What it does

  • Inspects USD stages to identify performance bottlenecks and structural inefficiencies.
  • Classifies assets into atomic, package, or aggregate structures to guide optimization.
  • Performs native USD optimization operations on disjoint scopes.
  • Publishes semantic libraries to enable efficient asset reuse.
  • Validates composition, ownership, and interface accessibility of the optimized output.
  • Profiles the final output against the original baseline to report measured improvements.

When to use it

  • When a USD stage suffers from slow loading times or high memory consumption.
  • When rendering performance (FPS) needs improvement in a production environment.
  • Before finalizing a USD assembly for downstream consumption.
  • When you need to standardize asset structures according to factory-level guidelines.

Advantages

  • Maintains asset boundaries and critical metadata during optimization.
  • Reduces scene complexity through intelligent reuse of shared definitions.
  • Provides verifiable reports on performance gains and composition integrity.
  • Supports large-scale scene optimization by processing independent definition scopes.

Usage examples

“Inspect the USD stage at /data/factory.usdc for performance issues and prepare a review workspace.”
The agent initializes the workspace, runs the inspection script, and provides a summary of candidates for optimization.
“Submit the optimization decisions for the factory stage based on the generated JSON.”
The agent processes the boundary decisions, executes the optimization operations, and generates the final optimized USD assembly.

Install

  1. Download the skill as a ZIP:

    https://autogineer.com/agent-skills/nvidia/omniverse-usd-performance-tuning/download
  2. Unzip it into the cross-client skills folder of your project:

    mkdir -p .agents/skills && unzip omniverse-usd-performance-tuning.zip -d .agents/skills

Documentation by NVIDIA Omniverse

Omniverse USD Performance Tuning

Purpose

Improve loading, memory use or rendering for the requested workload. Preserve the prim paths, appearance, placement and behavior required by downstream applications. Follow the component/subcomponent model in NVIDIA's Factory-Level USD Structuring guide. Apply the relevant parts of that model; a static asset does not need new animation or physics.

Inputs

Required: the source USD stage. When no workload is specified, use factory layout planning and visualization and record that assumption in the request/report. Explicit workload and preservation requirements take precedence. Optional: an output workspace, preservation requirements, physical error budget, existing library indexes and recorded runtime evidence. Use values from the current user request; reuse recorded values only while they remain applicable. Follow setup when no output path is supplied.

Prerequisites

Verify the selected runtime and required capabilities using setup. Keep its preflight record with the run; installation and remote access are conditional steps described there.

Instructions

  1. Inspect. Establish the workload, preservation requirements and output workspace. Set up progress and dashboard access before assessment. Verify setup and inspect the selected source configuration.

  2. Judge boundaries. Review equipment, service-system sections and architectural elements in the selected configuration, including animation and interiors. Record owners, editable interfaces, equipment categories and optional families using boundary guidance.

  3. Classify and author. Classify accepted assets children before parents. Capture or reuse the original profile before authoring. Prepare reusable definitions, materials and occurrence opinions using asset authoring.

  4. Optimize unique definitions. Run supported native operations concurrently on disjoint scopes with the recorded physical budget.

  5. Publish semantic libraries. Use one writer per category layer and remap consumer references.

  6. Check composition. Check authored ownership, references, bindings, placement, animation and interface accessibility through the written-output checks.

  7. Profile and report. Compare the final consumer with the original baseline under comparable conditions and report measured benefits, completion and operations by asset definition.

sequenceDiagram
    participant A as Agent
    participant T as USD tools
    participant W as Definition workers
    participant L as Library publisher
    participant D as Read-only progress dashboard
    Note over A,L: One compact progress state also renders text progress
    T-->>D: Recorded phase, assessment rounds, counters and decision reasons
    A->>T: Inspect original selected configuration
    T->>T: Collect cached child-to-parent summaries and repetition nominations
    T->>T: Load supplied indexes once and batch authored product clues
    T-->>A: Product review groups, member context, animation and reuse evidence
    loop Until boundaries and interiors are reviewed
        A->>A: Select asset boundaries, then identify required internal prims and properties
        A->>T: Submit boundary decisions with reviewed scope, required edits, category and optional family
        T->>T: Reuse unchanged source evidence, recompute boundary-dependent results
        opt Geometry ownership requires reconstruction
            T->>T: Regroup or native split, remap paths and reassess affected regions
        end
        T-->>A: Remaining candidates and member exceptions
    end
    T->>T: Classify selected child assets before parent assets, confirm reuse
    T->>T: Capture or reuse original baseline profile
    T->>T: Confirm accepted product identities from source and supplied indexes; reconcile complete parent-child groups
    T->>T: Skip worker inputs for confirmed groups; author remaining materials, editable prims and occurrence overrides
    T->>T: Determine merge groups on prepared interiors; preserve selected child assets and required prims
    T->>W: Read-only input USD layers for existing-library matching
    par Independent preparation jobs
        W->>W: Confirm remaining product, exact source or authorized normalized reuse; preserve occurrence placement
        W->>W: Consolidate accepted product or compatible exact/normalized leaf inputs; retain every occurrence
    end
    W-->>T: Confirmed library hits and definitions requiring optimization
    T->>W: Owned geometry prim paths in read-only input layers
    par Independent definition scopes
        W->>W: Remove small meshes when authorized; apply asset representation: reserve merge inputs or fit analytic prims; reduce, merge (native clustering if required), prune, flatten, prune
    end
    W->>L: Geometry outputs and assembly authoring outcomes
    L->>L: Share eligible static whole definitions within current category publication
    L->>L: Share remaining whole meshes; write USD reuse metadata and compact index
    L->>T: Output USD stage with rewritten references and dependencies
    T->>T: Check authored composition, ownership, animation and interfaces
    T->>T: Profile final output USD stage and report results by asset definition
    T-->>A: Measured changes, coverage and unresolved work
    T-->>D: Final verdict and report/library paths

Use the provided tools for traversal, native operations, authoring and checks; the agent supplies engineering judgments. Preserve current variant selections, authored animation, required materials and interfaces. Trust supported USD Optimize operations with valid settings and scopes. Deep geometric comparisons and whole-scene edit trials belong in development tests, not ordinary execution.

Complete when every unique definition has an executed or evidenced unchanged outcome, every asset has justified ownership and structure, and the written consumer passes composition checks. Retain native failures, unresolved judgments and missing required capabilities as incomplete work. Repeat only affected work when evidence changes.

Scope and authorization

  • Diagnosis and validation requests remain read-only. An optimization request authorizes work in a separate output workspace within its stated preservation limits.

  • Never overwrite source layers or inherited libraries without explicit authorization.

  • Reuse authorization already given. Ask only for a missing decision that changes identity, required appearance or a loss budget; explain the affected component and consequence.

  • Report only observed changes. Measure rendering before claiming higher FPS or lower GPU use.

Load only the relevant guidance

RequestRead
Validation onlySetup, then the validation section of operations
Approved operation chainSetup, operations, verification
Existing or persistent asset libraryAdd the library index requirements to the assessment
Missing standalone runtimeInstallation
Remote stageRemote access before the first open
Report from completed evidenceVerification

Use upstream package guidance for selected operation mechanics and proxy workflows. Keep large logs on disk as specified in setup.

Limitations

Convert CAD input with the source converter's workflow and NVIDIA's CAD conversion guidance. Render profiling requires NVIDIA/omniperf or the user's available profiler; USD open and traversal timings alone do not establish rendering improvements.

Troubleshooting

For unavailable imports, operations or remote assets, follow setup. For incomplete execution or failed composition checks, follow verification; retain the failed artifacts and report the missing capability or failed check.

Examples

After runtime preflight, use the verified Python interpreter from setup to inspect a stage:

python scripts/successor_assessment.py inspect --source /data/factory.usdc --workspace /data/factory-review

Review the generated candidates using boundary guidance, fill the generated decision template, then submit it:

python scripts/successor_assessment.py decide --workspace /data/factory-review --decisions /data/factory-review/decisions.json

These commands perform assessment; continue through the instructions above only when review is complete. Run from the skill directory; generated data stays in the workspace.

Output Format

Maintain live progress and editable steering throughout execution.

For end-to-end optimization, provide the written USD assembly, validated report JSON, and generated Markdown and HTML reports. State the runtime, operations actually executed, preservation and coverage results, measured changes, source-write status, and artifact paths. Diagnosis-only work may return a findings summary stating that no optimized stage was written.

Files

references/report-templates/render_preview.py

#!/usr/bin/env python3
# SPDX-FileCopyrightText: Copyright (c) 2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
# SPDX-License-Identifier: Apache-2.0
"""Render a USD Performance Tuning Report template.

The renderer intentionally supports only the small Jinja-compatible subset used
by the report templates: variable interpolation, for loops, if/else blocks, and
simple equality checks. It has no third-party runtime dependencies.

Block tags that sit alone on a line are stripped of that line's leading
whitespace and its trailing newline, matching Jinja's ``lstrip_blocks`` +
``trim_blocks``. Without that, every ``{% for %}`` iteration carried the newline
that followed the tag, so each rendered table row was separated by a blank line —
which ends the table in Markdown, turning the report's tables into loose runs of
pipe-delimited text.
"""
from __future__ import annotations

import argparse
import html
import json
import re
from collections.abc import Mapping
from pathlib import Path
from typing import Any


TEMPLATE_DIR = Path(__file__).resolve().parent
DEFAULT_TEMPLATE = TEMPLATE_DIR / "optimization-report.html.template"
TAG_RE = re.compile(r"{%\s*(.*?)\s*%}", re.DOTALL)
VAR_RE = re.compile(r"{{\s*(.*?)\s*}}", re.DOTALL)
#: A block tag alone on its line: nothing but whitespace before it, nothing but
#: whitespace and the newline after it. A tag sharing a line with content (the
#: HTML template's inline ``{% if group.caveat %}…{% endif %}``) does not match
#: and keeps its surrounding text exactly.
STANDALONE_TAG_RE = re.compile(r"^[ \t]*(\{%[^\n]*?%\})[ \t]*\n", re.MULTILINE)


def _score_percent(score: object) -> int:
    if not isinstance(score, (int, float)):
        return 0
    return max(0, min(100, round(float(score) * 10)))


def _score_display(score: object) -> str:
    if not isinstance(score, (int, float)):
        return "N/A"
    return f"{float(score):.1f}/10"


def _flag(value: object) -> str:
    if value is True:
        return "yes"
    if value is False:
        return "no"
    return "not recorded"


def _count(value: object) -> str:
    return f"{value:,}" if isinstance(value, int) and not isinstance(value, bool) else "not recorded"


def _gate_summary_rows(report: dict) -> list[dict]:
    """One line each for the machine-readable gate blocks the report carries.

    ``footprint`` used to be the only block with a rendered section, so a reader
    saw the disk numbers and none of the correctness state. These rows put the
    preservation, coverage, reuse, and fidelity gates in the deliverable without
    the agent having to restate them in free text. ``safety_gate`` gets its own
    section rather than a row — it is the one that can invalidate the run.
    """
    rows: list[dict] = []

    pres = report.get("preservation")
    if isinstance(pres, dict):
        rows.append({
            "name": "Preservation (silent loss)",
            "value": (
                f"{_count(pres.get('rendered_mesh_count'))} rendered meshes; "
                f"{_count(pres.get('dangling'))} dangling bindings; "
                f"distinct geometry bytes preserved: {_flag(pres.get('distinct_geometry_bytes_preserved'))}; "
                f"bounds preserved: {_flag(pres.get('bounds_preserved'))}"
            ),
        })

    coverage = report.get("target_coverage")
    if isinstance(coverage, dict):
        entries = [e for e in (coverage.get("entries") or []) if isinstance(e, Mapping)]
        tally: dict[str, int] = {}
        for entry in entries:
            tally[str(entry.get("disposition", "unrecorded"))] = (
                tally.get(str(entry.get("disposition", "unrecorded")), 0) + 1
            )
        breakdown = ", ".join(f"{name} {count}" for name, count in sorted(tally.items()))
        rows.append({
            "name": "Optimization target coverage",
            "value": (
                f"{len(entries)} target(s); ledger complete: {_flag(coverage.get('complete'))}"
                + (f" ({breakdown})" if breakdown else "")
            ),
        })

    structural = report.get("structural_summary")
    if isinstance(structural, dict):
        rows.append({
            "name": "Reuse and descent",
            "value": (
                f"reuse measured: {_flag(structural.get('reuse_measured'))}; "
                f"{_count(structural.get('unique_prototype_count'))} prototypes; "
                f"descent converged: {_flag(structural.get('frontier_descent_converged'))}"
            ),
        })

    fidelity = report.get("fidelity")
    if isinstance(fidelity, dict):
        band = fidelity.get("band")
        rows.append({
            "name": "Fidelity",
            "value": (
                f"max deviation within band: {_flag(fidelity.get('max_deviation_within_band'))}"
                + (f" (band {band})" if band else "")
            ),
        })

    return rows


def _quantity(value, unit=""):
    if not isinstance(value, (int, float)):
        return "Not recorded"
    if unit == "bytes":
        return f"{value / 1_000_000:,.1f} MB" if abs(value) >= 1_000_000 else f"{value:,.0f} bytes"
    if unit == "ms":
        return f"{value / 1000:,.2f} s" if abs(value) >= 1000 else f"{value:,.1f} ms"
    if unit == "seconds":
        return f"{value:,.2f} s"
    number = f"{value:,.0f}" if float(value).is_integer() else f"{value:,.1f}"
    return number + (" " + unit if unit else "")


def _asset_context(asset, names):
    row = dict(asset)
    row['search_text'] = ' '.join([asset.get('name', ''), asset.get('semantic_category', ''), asset.get('equipment_family', ''), asset.get('source_prim_path', ''), *asset.get('occurrences', [])])
    if row.get('structure_reason') == row.get('boundary_reason'):
        row['structure_reason'] = ''
    row['occurrence_count'] = len(asset.get('occurrences', []))
    row['child_names'] = [names.get(identifier, identifier) for identifier in asset.get('children', [])]
    row['metrics'] = [{'name': name.replace('_', ' ').title(),
                      'before': _quantity(asset.get('before', {}).get(name)),
                      'after': _quantity(asset.get('after', {}).get(name))}
                     for name in ('meshes', 'vertices', 'faces', 'analytic_primitives')]
    operations = [dict(op, seconds_display=_quantity(op.get('seconds'), 'seconds')) for op in asset.get('operations', [])]
    timings = [op['seconds'] for op in operations if isinstance(op.get('seconds'), (int, float))]
    row['operation_seconds'] = (_quantity(sum(timings), 'seconds') if timings and len(timings) == len(operations)
                                else (f"{_quantity(sum(timings), 'seconds')} recorded; {len(timings)}/{len(operations)} operations timed" if timings else 'Not recorded'))
    if not operations and asset.get('reused_from_library'):
        row['operation_seconds'] = '0.00 s (no native operations in this run)'
    row['prior_operations'] = [dict(op, seconds_display=_quantity(op.get('seconds'), 'seconds')) for op in asset.get('prior_operations', [])]
    row['operations'] = operations
    # Consecutive scopes of the same operation collapse without changing execution order.
    groups = []
    for operation in operations:
        if not groups or groups[-1]['name'] != operation['name']:
            groups.append({'name': operation['name'], 'count': 0, 'statuses': {}})
        group = groups[-1]
        group['count'] += 1
        status = operation['status'].replace('_', ' ')
        group['statuses'][status] = group['statuses'].get(status, 0) + 1
    row['operation_groups'] = [dict(name=group['name'], outcomes=', '.join(f"{count} {status}" for status, count in group['statuses'].items())) for group in groups]
    return row


def build_context(report: dict) -> dict:
    context = dict(report)
    score = report.get("optimization_score")
    context["score_degrees"] = round(float(score) * 36, 1) if isinstance(score, (int, float)) else 0
    context.setdefault("executive_summary", "")
    verdict = report.get('verdict', 'neutral')
    context['verdict_label'] = {'improved': 'Improved', 'mixed': 'Mixed', 'neutral': 'Unchanged', 'regressed': 'Regressed'}.get(verdict, 'Unchanged')
    context['verdict_color'] = {'improved': 'var(--good)', 'mixed': 'var(--warn)', 'neutral': 'var(--unknown)', 'regressed': 'var(--bad)'}.get(verdict, 'var(--unknown)')
    groups = report.get('metric_groups', [])
    insufficient = (report.get('measurement_context', {}).get('storage_coverage') == 'partial'
                    or not report.get('metrics')
                    or (bool(groups) and all(g.get('status') == 'not_measured' for g in groups)))
    if verdict == 'neutral' and insufficient:
        context['verdict_label'] = 'Insufficient measurements'
        context['verdict_color'] = 'var(--unknown)'
    context['completion'] = report.get('report_summary', {}).get('completion',
        'Recorded target coverage complete.' if report.get('target_coverage', {}).get('complete') else 'Target coverage incomplete or not recorded.')
    context['note_paragraphs'] = [p for p in report.get('notes', '').split('\n\n') if p]
    names = {a['definition_id']: a['name'] for a in report.get('asset_results', [])}
    context['assets'] = [_asset_context(a, names) for a in report.get('asset_results', [])]
    # The legacy flat operation list remains available for old reports. Successor shared operations follow asset receipts.
    context['shared_operations'] = [op for op in report.get('operations', []) if op.get('name') in ('optimizeMaterials', 'repairMaterialTargets')] if context['assets'] else []
    context['execution_summary'] = report.get('measurement_context', {}).get('execution_counts', {})
    context['next_run_text'] = '\n'.join(report.get('report_summary', {}).get('next_run_instructions', []))
    context.setdefault("reasoning", "")
    context["reasoning_paragraphs"] = [
        paragraph.strip()
        for paragraph in str(context["reasoning"]).split("\n\n")
        if paragraph.strip()
    ]

    groups = []
    for group in report.get("metric_groups", []):
        item = dict(group)
        item["score_percent"] = _score_percent(item.get("score"))
        item["score_display"] = _score_display(item.get("score"))
        groups.append(item)
    context["metric_groups"] = groups

    metrics = []
    for metric in report.get("metrics", []):
        item = dict(metric)
        item.setdefault("display_name", item.get("name", ""))
        item.setdefault("unit", "")
        item.setdefault("evidence_type", "direct")
        unit = item['unit'] or ('bytes' if item.get('name', '').endswith('_bytes') else '')
        item['before_display'] = _quantity(item.get('before'), unit)
        item['after_display'] = _quantity(item.get('after'), unit)
        delta = item.get('change_pct')
        item['change_display'] = (('Added ' + item['after_display']) if item.get('before') == 0 and item.get('after', 0) > 0 else
                                  f"{delta:+.1f}%" if isinstance(delta, (int, float)) else 'Not measured')
        metrics.append(item)
    context["metrics"] = metrics
    context['headline_metrics'] = [m for name in ('meshes', 'vertices', 'faces', 'warm_open_ms', 'referenced_storage_bytes') for m in metrics if m.get('name') == name]
    footprint = dict(report.get('footprint') or {})
    for key, value in list(footprint.items()):
        if key.endswith('_bytes'):
            footprint[key] = _quantity(value, 'bytes')
        elif key.endswith('_pct') and isinstance(value, (int, float)):
            footprint[key] = f"{value:+.1f}"
    context['footprint'] = footprint

    measurement_context = report.get("measurement_context", {})
    context["measurement_context_items"] = [
        {"name": key.replace("_", " ").title(), "value": "N/A" if value is None else value}
        for key, value in measurement_context.items() if key != "execution_counts"
    ]
    runtime_profiling = report.get("runtime_profiling", {})
    context["runtime_profiling_items"] = [
        {"name": key.replace("_", " ").title(), "value": "N/A" if value is None else value}
        for key, value in runtime_profiling.items()
    ]
    context["gate_summary_rows"] = _gate_summary_rows(report)
    return context


def should_autoescape(template_name: str | None) -> bool:
    return bool(template_name and template_name.endswith(".html.template"))


def _resolve(name: str, context: Mapping[str, Any]) -> Any:
    name = name.strip()
    if not name:
        return ""
    if (name.startswith('"') and name.endswith('"')) or (name.startswith("'") and name.endswith("'")):
        return name[1:-1]
    value: Any = context
    for part in name.split("."):
        if isinstance(value, Mapping):
            value = value.get(part, "")
        else:
            value = getattr(value, part, "")
    return value


def _eval_condition(expression: str, context: Mapping[str, Any]) -> bool:
    expression = expression.strip()
    if expression.startswith("not "):
        return not _eval_condition(expression[4:], context)
    for operator in ("==", "!="):
        if operator in expression:
            left, right = expression.split(operator, 1)
            result = _resolve(left, context) == _resolve(right, context)
            return result if operator == "==" else not result
    return bool(_resolve(expression, context))


def _find_endfor(template: str, pos: int) -> tuple[int, int]:
    depth = 1
    for match in TAG_RE.finditer(template, pos):
        command = match.group(1).strip()
        if command.startswith("for "):
            depth += 1
        elif command == "endfor":
            depth -= 1
            if depth == 0:
                return match.start(), match.end()
    raise ValueError("missing {% endfor %} in report template")


def _find_if_parts(template: str, pos: int) -> tuple[int | None, int | None, int, int]:
    depth = 1
    else_start: int | None = None
    else_end: int | None = None
    for match in TAG_RE.finditer(template, pos):
        command = match.group(1).strip()
        if command.startswith("if "):
            depth += 1
        elif command == "else" and depth == 1:
            else_start = match.start()
            else_end = match.end()
        elif command == "endif":
            depth -= 1
            if depth == 0:
                return else_start, else_end, match.start(), match.end()
    raise ValueError("missing {% endif %} in report template")


def _render_variables(text: str, context: Mapping[str, Any], autoescape: bool) -> str:
    def replace(match: re.Match[str]) -> str:
        value = _resolve(match.group(1), context)
        if value is None:
            value = "N/A"
        rendered = str(value)
        return html.escape(rendered, quote=True) if autoescape else rendered

    return VAR_RE.sub(replace, text)


def _render_block(template: str, context: Mapping[str, Any], autoescape: bool) -> str:
    rendered: list[str] = []
    pos = 0

    while True:
        match = TAG_RE.search(template, pos)
        if not match:
            rendered.append(_render_variables(template[pos:], context, autoescape))
            break

        rendered.append(_render_variables(template[pos:match.start()], context, autoescape))
        command = match.group(1).strip()

        if command.startswith("for "):
            loop_target, _, expression = command[4:].partition(" in ")
            if not loop_target or not expression:
                raise ValueError(f"unsupported for expression: {command}")
            end_start, end_end = _find_endfor(template, match.end())
            body = template[match.end():end_start]
            for item in _resolve(expression, context) or []:
                child_context = dict(context)
                child_context[loop_target.strip()] = item
                rendered.append(_render_block(body, child_context, autoescape))
            pos = end_end
            continue

        if command.startswith("if "):
            else_start, else_end, endif_start, endif_end = _find_if_parts(template, match.end())
            if else_start is None:
                true_body = template[match.end():endif_start]
                false_body = ""
            else:
                true_body = template[match.end():else_start]
                false_body = template[else_end:endif_start]
            rendered.append(
                _render_block(
                    true_body if _eval_condition(command[3:], context) else false_body,
                    context,
                    autoescape,
                )
            )
            pos = endif_end
            continue

        if command in {"else", "endif", "endfor"}:
            raise ValueError(f"unexpected template tag: {command}")

        raise ValueError(f"unsupported template tag: {command}")

    return "".join(rendered)


def trim_block_tag_lines(template: str) -> str:
    """Collapse each standalone block tag onto the text around it.

    Jinja's ``lstrip_blocks`` + ``trim_blocks``. A ``{% for %}`` tag on its own
    line otherwise donates that line's newline to every iteration of the loop
    body, so an N-row table renders as N rows each preceded by a blank line —
    and a blank line ends a Markdown table.
    """
    return STANDALONE_TAG_RE.sub(lambda match: match.group(1), template)


def render_template(template: str, context: Mapping[str, Any], *, autoescape: bool) -> str:
    template = template.replace('/* shared-report-theme */', (TEMPLATE_DIR/'theme.css').read_text())
    return _render_block(trim_block_tag_lines(template), context, autoescape)


def main() -> int:
    parser = argparse.ArgumentParser(description=__doc__)
    parser.add_argument("--fixture", type=Path, required=True, help="Report JSON to render")
    parser.add_argument("--template", type=Path, default=DEFAULT_TEMPLATE)
    parser.add_argument("--output", type=Path, required=True, help="Output file in the run workspace")
    args = parser.parse_args()

    report = json.loads(args.fixture.read_text(encoding="utf-8"))
    rendered = render_template(
        args.template.read_text(encoding="utf-8"),
        build_context(report),
        autoescape=should_autoescape(args.template.name),
    )
    args.output.write_text(rendered, encoding="utf-8")
    print(args.output)
    return 0


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

Requirements and compatibility

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 Python, Needs Node.js

The author says: Requires a verified standalone Usd Optimize/OpenUSD runtime for operations, usd-validation-nvidia for supported checks, and a writable output workspace. Remote assets require a working resolver. Render profiling is an external handoff.

Security

What our scanner noticed· Medium risk

  • Large base64-encoded blobs
  • 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
NVIDIA Omniverse
Publisher
NVIDIA (Official)
Source
NVIDIA/skills · skills/omniverse-usd-performance-tuning · commit fce3cca
Licence
Apache License 2.0 (redistribution allowed)
  • LICENSE (added by AutoGineer from LICENSE-APACHE)
  • ATTRIBUTION.md (written by AutoGineer)
Synced

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

Download ZIP

The ZIP holds the skill’s files, its licence files and an ATTRIBUTION.md. Download omniverse-usd-performance-tuning.zip