NVIDIA
A list of OpenClaw skills by NVIDIA on aiagents.org.
392 listings · nvidia on GitHub
Claim this page Make Grok Bots, skills or prompts? Submit them.
accelerated-computing-cudfOpenClaw
NVIDIA's official guidance for GPU DataFrames with cuDF, including pandas acceleration and multi-GPU ETL.
amc-run-rtsp-calibrationOpenClaw
NVIDIA skill for calibrating cameras from live RTSP streams with AutoMagicCalib.
amc-run-sample-calibrationOpenClaw
NVIDIA skill for running a complete calibration on the bundled sample dataset with AMC.
amc-run-video-calibrationOpenClaw
NVIDIA skill for calibrating pre-recorded MP4 camera datasets with AutoMagicCalib.
amc-setup-calibration-stackOpenClaw
NVIDIA skill for launching the AutoMagicCalib service and its browser interface with Docker Compose.
CAD to SimReadyOpenClaw
NVIDIA skill for running the Omniverse pipeline from source asset to SimReady asset.
cudaq-guideOpenClaw
Guide to setting up CUDA-Q and writing quantum kernels for simulators and QPUs.
cudaq-importingOpenClaw
Ports circuits from other frameworks like Qiskit to CUDA-Q kernels, preserving results.
cuopt-developerOpenClaw
NVIDIA skill for modifying, building, testing, and contributing to cuOpt.
cuopt-installOpenClaw
NVIDIA skill for installing and verifying cuOpt through pip, conda, or Docker.
cuopt-multi-objective-explorationOpenClaw
NVIDIA skill for tracing Pareto frontiers across competing objectives with repeated cuOpt solves.
cuopt-numerical-optimization-apiOpenClaw
NVIDIA skill for solving LP, MILP, and QP with cuOpt in Python, C, or CLI.
cuopt-numerical-optimization-formulationOpenClaw
NVIDIA skill for formulating LP, MILP, and QP problems from text, without API details.
cuopt-routing-api-pythonOpenClaw
NVIDIA skill for solving vehicle routing problems with the cuOpt Python API.
cuopt-server-api-pythonOpenClaw
NVIDIA skill for running the cuOpt REST server and calling it from Python or curl.
DALI Dynamic ModeOpenClaw
NVIDIA skill for writing and migrating code with the imperative DALI API.
data-designerOpenClaw
NVIDIA skill for creating datasets, generating synthetic data, and building generation pipelines.
DeepStream Import Vision ModelOpenClaw
NVIDIA skill for deploying object detection models into DeepStream.
deepstream-devOpenClaw
NVIDIA skill for DeepStream SDK development with the Python pyservicemaker API.
deepstream-generate-pipelineOpenClaw
NVIDIA skill for interactively building DeepStream GStreamer pipelines for inference and streaming.
deepstream-profile-pipelineOpenClaw
NVIDIA skill for profiling DeepStream pipelines with Nsight Systems and deriving configs.
deepstream-run-mv3dtOpenClaw
NVIDIA skill for running MV3DT, DeepStream's reference app for tracking in 3D across cameras.
deepstream-sopOpenClaw
NVIDIA skill for building and debugging the DeepStream SOP microservice that checks assembly steps.
dicom-metadata-extractOpenClaw
NVIDIA skill for extracting chosen metadata from a DICOM file and flagging PHI tags.
dicom-series-preflightOpenClaw
NVIDIA skill for a headers-only preflight of a DICOM series before conversion or inference.
dicom-series-to-volumeOpenClaw
NVIDIA skill for converting a CT DICOM series into a Hounsfield-unit NIfTI volume.
diffdock-nimOpenClaw
NVIDIA skill for predicting small-molecule binding poses with DiffDock through NIM.
digital-health-clinical-asr-buildOpenClaw
NVIDIA skill for the second Clinical ASR Flywheel stage: clinical terms, IPA tags, NeMo manifest.
digital-health-clinical-asr-evalOpenClaw
NVIDIA skill for scoring a NeMo manifest into a KER leaderboard for clinical ASR.
digital-health-clinical-asr-finetuneOpenClaw
NVIDIA skill for fine-tuning Parakeet on clinical terms when priority KER stays high.
digital-health-clinical-asr-setupOpenClaw
NVIDIA skill for starting a Clinical ASR Flywheel cycle with checks and smoke tests.
doca-aes-gcmOpenClaw
NVIDIA skill for DOCA AES-GCM encrypt and decrypt tasks on BlueField and ConnectX.
doca-argpOpenClaw
NVIDIA skill for adding and managing command-line flags with the DOCA Arg Parser.
doca-argusOpenClaw
NVIDIA skill for deploying and operating the DOCA Argus runtime-security service on BlueField.
doca-bare-metal-deploymentOpenClaw
NVIDIA skill for running and recovering DOCA binaries directly on BlueField hardware.
doca-benchOpenClaw
NVIDIA skill for measuring DOCA throughput and latency with the doca_bench tool.
doca-bench-extensionOpenClaw
NVIDIA skill for writing and debugging custom doca-bench plug-ins.
doca-bf3-deploymentOpenClaw
NVIDIA skill for BlueField-3 first-day bring-up over RShim, including pushing a BFB bundle.
doca-bf4-deploymentOpenClaw
NVIDIA skill warning about and guiding risky BlueField-4 hardware operations.
doca-capsOpenClaw
NVIDIA skill for checking what DOCA detects on a host with the doca_caps tool.
doca-collectx-deploymentOpenClaw
NVIDIA skill for deploying a CollectX-based DOCA telemetry collector with its exporters.
doca-comchOpenClaw
NVIDIA skill for DOCA Comch messaging between a host and its BlueField over PCIe.
doca-commonOpenClaw
NVIDIA skill for the foundation DOCA programming primitives shared by every library context.
doca-compressOpenClaw
NVIDIA skill for DOCA Compress programming: deflate and LZ4 tasks on BlueField and ConnectX.
doca-container-deploymentOpenClaw
NVIDIA skill for deploying in-bundle DOCA service containers on BlueField with kubelet.
doca-debugOpenClaw
NVIDIA skill for debugging DOCA build, link, and runtime problems with a layered ladder.
doca-devemuOpenClaw
NVIDIA skill for DOCA Device Emulation: exposing custom emulated PCIe devices from a BlueField.
doca-dmaOpenClaw
NVIDIA skill for DOCA DMA programming: memcpy tasks, buffer sizing, and permissions.
doca-dmsOpenClaw
NVIDIA skill for operating the DOCA Management Service on BlueField, hosts, or Kubernetes pods.
doca-dpaOpenClaw
NVIDIA skill for host-side DOCA DPA programming: contexts, app images, threads, and kernel launches.
doca-dpa-hl-tracerOpenClaw
NVIDIA skill for capturing and decoding DPA programming-event traces with doca_dpa_hl_tracer.
doca-dpdk-bridgeOpenClaw
NVIDIA skill for adding DOCA capabilities such as Flow to an existing DPDK application.
doca-erasure-codingOpenClaw
NVIDIA skill for DOCA Erasure Coding programming: create, recover, and update tasks.
doca-ethOpenClaw
NVIDIA skill for DOCA Ethernet packet-queue programming with RX and TX queues.
doca-fireflyOpenClaw
NVIDIA skill for operating the DOCA Firefly PTP time-sync service on BlueField.
doca-flowOpenClaw
NVIDIA skill for building and debugging DOCA Flow apps with match-action pipes.
doca-flow-dpa-perfOpenClaw
NVIDIA skill for measuring rule update and disable rates on DOCA Flow with DPA offload.
doca-flow-dpa-providerOpenClaw
NVIDIA skill for exporting DOCA Flow pipes into DPA address space for DPACC kernels.
doca-flow-grpc-serverOpenClaw
NVIDIA skill noting the DOCA Flow gRPC server is plaintext-only and needs outside transport security.
doca-flow-perfOpenClaw
NVIDIA skill for measuring DOCA Flow pipeline control-plane rates with doca_flow_perf.
doca-flow-tuneOpenClaw
NVIDIA skill for tuning running or captured DOCA Flow pipelines with doca_flow_tune.
doca-gpiOpenClaw
NVIDIA skill for DOCA GPI programming so CUDA kernels drive RDMA queues from GPU memory.
doca-gpunetioOpenClaw
NVIDIA skill for DOCA GPUNetIO programming that connects CUDA kernels to Ethernet queues.
doca-gpunetio-ib-write-bwOpenClaw
NVIDIA skill for building and reading the GPUNetIO RDMA WRITE bandwidth benchmark.
doca-gpunetio-ib-write-latOpenClaw
NVIDIA skill for measuring GPU-initiated RDMA WRITE latency with GPUNetIO.
doca-hardware-safetyOpenClaw
NVIDIA skill for gating hardware-changing DPU and NIC operations behind safety checks.
doca-mgmtOpenClaw
NVIDIA skill for programming DOCA Management APIs on BlueField and ConnectX.
doca-pccOpenClaw
NVIDIA skill for loading a custom congestion control algorithm onto BlueField from the host.
doca-pcc-countersOpenClaw
NVIDIA skill for arming and reading DOCA PCC diagnostic counters with pcc_counters.sh.
doca-pcc-ztr-rttcc-algoOpenClaw
NVIDIA skill for deploying and tuning the Zero-Touch RTT-based congestion control algorithm on BlueField-3.
doca-programming-guideOpenClaw
NVIDIA skill for first DOCA apps and library-agnostic programming questions.
doca-public-knowledge-mapOpenClaw
NVIDIA skill for finding authoritative DOCA documentation when the source tree is unavailable.
doca-rdmaOpenClaw
NVIDIA skill for programming DOCA RDMA contexts, connection methods, and task types.
doca-rdmiOpenClaw
NVIDIA skill for DOCA RDMI programming of one-sided RDMA flows started by accelerators.
doca-rmaxOpenClaw
NVIDIA skill for DOCA Rivermax receive streams that carry precisely timed media over IP.
doca-setupOpenClaw
NVIDIA skill for preparing and verifying the DOCA environment around a workload.
doca-shaOpenClaw
NVIDIA skill for offloading SHA hashing to BlueField and ConnectX with DOCA SHA.
doca-sha-offload-engineOpenClaw
NVIDIA skill for plugging DOCA's SHA offload into OpenSSL pipelines through an engine.
doca-socket-relayOpenClaw
NVIDIA skill for bridging socket-based host apps to a BlueField with DOCA Socket Relay.
doca-spcx-ccOpenClaw
NVIDIA skill for running congestion control algorithms on BlueField DPAs with doca_spcx_cc.
doca-staOpenClaw
NVIDIA skill for building NVMe-over-Fabrics storage targets with DOCA STA on BlueField.
doca-structured-tools-contractOpenClaw
NVIDIA skill for consolidated one-shot DOCA answers through the structured tools contract.
doca-telemetryOpenClaw
NVIDIA skill for reading DOCA hardware-counter events through per-domain telemetry libraries.
doca-telemetry-exporterOpenClaw
NVIDIA skill for programming the DOCA Telemetry Exporter with schemas, event types, and publish surfaces.
doca-telemetry-utilsOpenClaw
NVIDIA skill for using doca_telemetry_utils to look up diagnostic counter schemas and support.
doca-upgradeOpenClaw
NVIDIA skill for guided DOCA upgrades and downgrades that detect, report, and ask first.
doca-uromOpenClaw
NVIDIA skill for host-side DOCA UROM work that offloads remote memory operations to BlueField.
doca-urom-svcOpenClaw
NVIDIA skill for running the UROM service container for DOCA on BlueField's Arm cores.
doca-verbsOpenClaw
NVIDIA skill for low-level DOCA verbs: managing QPs, CQs, and memory regions below higher-level libraries.
doca-versionOpenClaw
NVIDIA skill for detecting and validating DOCA versions across install sources and containers.
dynamo-interconnect-checkOpenClaw
NVIDIA skill for verifying the NIXL, UCX, and NCCL interconnect of a Dynamo deployment.
dynamo-recipe-runnerOpenClaw
NVIDIA skill for selecting, validating, patching, and deploying Dynamo Kubernetes recipes.
dynamo-router-starterOpenClaw
NVIDIA skill for starting or patching Dynamo router modes and smoke-testing router endpoints.
dynamo-troubleshootOpenClaw
NVIDIA skill for diagnosing Dynamo deployments that fail or run unhealthy.
earth2studio-create-datasourceOpenClaw
NVIDIA skill for creating and validating Earth2Studio data source wrappers around remote stores.
earth2studio-create-diagnosticOpenClaw
NVIDIA skill for building Earth2Studio wrappers for models that transform data in one step.
earth2studio-create-prognosticOpenClaw
NVIDIA skill for creating Earth2Studio prognostic forecast model wrappers.
earth2studio-data-fetchOpenClaw
NVIDIA skill for fetching weather and climate data for chosen variables and times via Earth2Studio.
earth2studio-deterministic-forecastOpenClaw
NVIDIA skill for writing single-run forecast scripts with Earth2Studio.
earth2studio-discoverOpenClaw
NVIDIA skill for finding Earth2Studio models, data, and sample code for weather or climate tasks.
earth2studio-installOpenClaw
NVIDIA skill for installing Earth2Studio with uv or pip and configuring the environment.
evo2-nimOpenClaw
NVIDIA skill for generating and analyzing DNA sequences with the Evo 2 NIM.
FoundationPose PipelineOpenClaw
NVIDIA skill for running BOP datasets and assessing FoundationPose pose metrics.
FoundationPose SetupOpenClaw
NVIDIA skill for installing and repairing the FoundationPose perception pipeline.
g-assist-mcp-skillOpenClaw
NVIDIA skill for checking and changing display and GPU settings on this machine.
genmol-nimOpenClaw
NVIDIA skill for generating drug-like molecules with the GenMol NIM.
holohub-app-lifecycleOpenClaw
NVIDIA skill for everyday HoloHub app work with the holohub command: scaffold, build, run, test.
holohub-debug-build-runOpenClaw
NVIDIA skill for diagnosing failing, hanging, or misbehaving HoloHub commands.
holohub-module-lifecycleOpenClaw
NVIDIA skill for Holoscan Module work: scaffolds, tests, editable installs, and packaging.
holoscan-install-condaOpenClaw
NVIDIA skill for installing Holoscan SDK 4.3 or later with Conda on CUDA 13.
holoscan-install-containerOpenClaw
NVIDIA skill for installing Holoscan SDK from the NGC Docker container.
holoscan-install-debianOpenClaw
NVIDIA skill for installing Holoscan SDK on Ubuntu natively through apt.
holoscan-install-sourceOpenClaw
NVIDIA skill for compiling Holoscan SDK from source with the run script in the tree.
holoscan-install-wheelOpenClaw
NVIDIA skill for installing the Holoscan SDK Python wheel with pip into a virtual environment.
holoscan-setupOpenClaw
NVIDIA skill for guiding Holoscan SDK installs by checking the host and picking a method.
hsb-appOpenClaw
NVIDIA skill for discovering and running Holoscan Sensor Bridge example apps on a devkit.
hsb-flashOpenClaw
NVIDIA skill for flashing FPGA firmware on HSB boards attached to an NVIDIA devkit.
hsb-ip-create-topOpenClaw
NVIDIA skill for creating or explaining HSB top-level FPGA wrappers from validated def files.
hsb-ip-defOpenClaw
NVIDIA skill for generating, validating, comparing, and explaining HSB hololink def macros.
hsb-ip-packetizerOpenClaw
NVIDIA skill for choosing or explaining HSB RX packetizer fields for sensors in def files.
hsb-setupOpenClaw
NVIDIA skill for setting up Holoscan Sensor Bridge from the repo through demo and verification.
hsb-testOpenClaw
NVIDIA skill for running QA test plans against Holoscan Sensor Bridge devices and reporting results.
i4h-lerobot-vizOpenClaw
NVIDIA skill for serving a converted LeRobot dataset and inspecting it visually in a browser.
i4h-workflowOpenClaw
NVIDIA skill explaining the runtime for i4h workflows and routing to stage skills.
i4h-workflow-createOpenClaw
NVIDIA skill for scaffolding a minimal blank i4h Workflow with a ground-and-light Scene.
i4h-workflow-dataset-annotateOpenClaw
NVIDIA skill for grading or filtering workflow HDF5 episodes with a vision model.
i4h-workflow-dataset-convertOpenClaw
NVIDIA skill for converting workflow HDF5 recordings into LeRobot datasets.
i4h-workflow-dataset-mimicOpenClaw
NVIDIA skill for multiplying workflow HDF5 demos by jittering actions into synthetic variants.
i4h-workflow-dataset-replayOpenClaw
NVIDIA skill for replaying a recorded HDF5 episode inside the Scene it came from.
i4h-workflow-dataset-teleopOpenClaw
NVIDIA skill for recording demonstrations into workflow HDF5 through a teleop Task.
i4h-workflow-e2eOpenClaw
NVIDIA skill for running the maintained data-to-policy workflow pipeline end to end.
i4h-workflow-finetuneOpenClaw
NVIDIA skill for fine-tuning GR00T or openpi remote Tasks on compatible LeRobot data.
i4h-workflow-scene-editOpenClaw
NVIDIA skill for editing the Scene or task contract of a workflow that already exists.
i4h-workflow-setupOpenClaw
NVIDIA skill for preflight checks and installation of the top-level i4h workflow runtime.
i4h-workflow-validateOpenClaw
NVIDIA skill for running policy or rule-driven rollouts and verifying simulator success.
isaac-mission-control-showcaseOpenClaw
NVIDIA skill for running and validating a full Mission Control showcase in Isaac Sim.
jetson-build-sourceOpenClaw
NVIDIA skill for rebuilding the Jetson BSP overlay from sources: device tree, modules, or kernel.
jetson-customize-cameraOpenClaw
NVIDIA skill for enabling MIPI and GMSL cameras on custom Jetson carriers using overlays.
jetson-customize-clocksOpenClaw
NVIDIA skill for locking or capping Jetson CPU, GPU, and EMC clocks before flashing.
jetson-customize-fanOpenClaw
NVIDIA skill for adding, editing, or removing nvfancontrol fan profiles on Jetson targets.
jetson-customize-mgbeOpenClaw
NVIDIA skill for enabling Jetson Thor multi-gigabit MGBE QSFP ports with a DT overlay.
jetson-customize-nvpmodelOpenClaw
NVIDIA skill for adding, editing, or removing nvpmodel power modes on Jetson targets.
jetson-customize-pcieOpenClaw
NVIDIA skill for enabling, disabling, and configuring PCIe controllers on custom Jetson carriers.
jetson-customize-pinmuxOpenClaw
NVIDIA skill for configuring per-pin function, direction, and initial state from the Jetson pinmux spreadsheet.
jetson-customize-uphyOpenClaw
NVIDIA skill for configuring Jetson UPHY lane allocation on custom Orin and Thor carriers.
jetson-customize-usbOpenClaw
NVIDIA skill for enabling or disabling Jetson USB2 and USB3 ports with a DT overlay.
jetson-derive-carrierOpenClaw
NVIDIA skill for bootstrapping a custom Jetson carrier board from the reference devkit files.
jetson-diagnosticOpenClaw
NVIDIA skill for taking a health snapshot of a Jetson without changing anything.
jetson-download-bspOpenClaw
NVIDIA skill for downloading Jetson Linux BSP artifacts for the active target.
jetson-flash-imageOpenClaw
NVIDIA skill for flashing a promoted BSP image onto a Jetson in recovery mode.
jetson-generate-kbOpenClaw
NVIDIA skill for generating a per-target markdown knowledge base from the BSP and source trees.
jetson-headless-modeOpenClaw
NVIDIA skill for planning and applying safe changes that put a Jetson into headless mode.
jetson-inference-mem-tuneOpenClaw
NVIDIA skill for choosing serving stacks and memory flags for LLMs on Jetson.
jetson-init-imageOpenClaw
NVIDIA skill for unpacking Jetson Linux and rootfs and applying binaries for the active target.
jetson-init-sourceOpenClaw
NVIDIA skill for preparing the BSP source workspace with its overlay tracker and cross toolchain.
jetson-init-targetOpenClaw
NVIDIA skill for writing a new Jetson platform profile and making it the active one.
jetson-link-docsOpenClaw
NVIDIA skill for binding already-downloaded Jetson reference docs into the active profile.
jetson-llm-benchmarkOpenClaw
NVIDIA skill for benchmarking LLM and VLM serving speed on Jetson across three runtimes.
jetson-llm-serveOpenClaw
NVIDIA skill for launching LLM serving with SGLang or vLLM on Jetson.
jetson-memory-auditOpenClaw
NVIDIA skill for measuring Jetson DRAM and NvMap usage and verifying memory reclamation.
jetson-optimize-memoryOpenClaw
NVIDIA skill to free Jetson DRAM by switching off unused subsystems.
jetson-packageOpenClaw
NVIDIA skill for picking Jetson-compatible containers, vLLM images, and AI Lab package indexes.
jetson-print-bsp-infoOpenClaw
NVIDIA example skill that prints BSP details from a host-side Tegra root.
jetson-print-device-infoOpenClaw
NVIDIA example skill that prints details of a running Jetson.
jetson-promote-imageOpenClaw
NVIDIA skill for moving overlay files and build output to the BSP image staging area.
jetson-quick-startOpenClaw
NVIDIA entry skill for Jetson and IGX BSP customization using one setup questionnaire.
jetson-set-targetOpenClaw
NVIDIA skill for switching the active Jetson platform to a profile that already exists.
jetson-speculative-decodingOpenClaw
NVIDIA skill for enabling speculative decoding with EAGLE-3 on a Jetson vLLM server.
jetson-validate-imageOpenClaw
NVIDIA skill for checking a flashed Jetson image with static and on-device tests.
jetson-video-benchmarkOpenClaw
NVIDIA skill for measuring Jetson video codec encode and decode throughput.
jetson-video-capabilityOpenClaw
NVIDIA skill for reconciling Jetson video codec support from live APIs, samples, and docs.
jetson-video-pipelineOpenClaw
NVIDIA skill for planning, running, and separately checking Jetson video codec workflows.
jetson-video-recipeOpenClaw
NVIDIA skill for turning a Jetson encoding scenario into one API-agnostic recipe.
jetson-video-setupOpenClaw
NVIDIA skill for installing, repairing, and verifying native video codec tooling on Jetson.
kermt-add-cmim-pretrainOpenClaw
NVIDIA skill for converting a KERMT checkpoint to a hybrid cMIM one and pretraining further.
kermt-continue-pretrainOpenClaw
NVIDIA skill for extending KERMT pretraining with your own SMILES data.
kermt-embedOpenClaw
NVIDIA skill for extracting per-molecule embeddings from KERMT encoder checkpoints.
kermt-finetuneOpenClaw
NVIDIA skill for training a pretrained KERMT encoder with your labeled CSV data.
kermt-inferOpenClaw
NVIDIA skill for predicting properties from SMILES with a fine-tuned KERMT checkpoint.
kermt-monitorOpenClaw
NVIDIA skill for checking progress of detached KERMT pretrain and finetune runs.
kermt-pretrain-scratchOpenClaw
NVIDIA skill for pretraining a brand-new KERMT model on a corpus you provide.
kermt-setupOpenClaw
NVIDIA skill for bootstrapping the KERMT environment with Docker, GPU checks, and the image.
launch-nemo-rlOpenClaw
NVIDIA playbook for running NeMo-RL on Kubernetes, from launch through debugging.
mcore-create-issueOpenClaw
NVIDIA skill for investigating a failed CI run and filing a GitHub issue for it.
mcore-linting-and-formattingOpenClaw
NVIDIA skill for linting and formatting Megatron-LM code with its standard tools.
mcore-run-on-slurmOpenClaw
NVIDIA guide for starting distributed Megatron-LM training on SLURM.
mcore-split-prOpenClaw
NVIDIA skill for splitting a large PR into smaller ones to cut CODEOWNERS review groups.
mcore-testingOpenClaw
NVIDIA skill for the Megatron-LM test system: layout, recipes, and unit and functional tests.
medtech-model-evidence-exportOpenClaw
NVIDIA skill for exporting sanitized evidence from medical AI inference to MLflow.
molmim-nimOpenClaw
NVIDIA skill for generating and optimizing small molecules in latent space with MolMIM.
msa-search-nimOpenClaw
NVIDIA skill for building protein multiple sequence alignments with ColabFold's NIM search service.
msa-structure-prediction-pipelineOpenClaw
NVIDIA skill for a full protein structure prediction pipeline from MSA search to folding.
nemo-automodel-distributed-trainingOpenClaw
NVIDIA guide for picking and setting up distributed training strategies for AutoModel.
nemo-automodel-launcher-configOpenClaw
NVIDIA skill for configuring how NeMo AutoModel jobs launch: interactive, Slurm, or SkyPilot.
nemo-automodel-model-onboardingOpenClaw
NVIDIA guide for adding new model architectures to AutoModel.
nemo-automodel-recipe-developmentOpenClaw
NVIDIA skill for creating and editing AutoModel recipes for training and evaluation.
nemo-fabric-integrateOpenClaw
NVIDIA skill for integrating NeMo Fabric into apps and platforms through its typed Python SDK.
nemo-mbridge-mlm-bridge-trainingOpenClaw
NVIDIA skill for running Megatron-LM and Bridge training on mock or real data.
nemo-mbridge-multi-node-slurmOpenClaw
NVIDIA skill for turning single-node scripts into multi-node Slurm jobs and debugging them.
nemo-mbridge-perf-activation-recomputeOpenClaw
NVIDIA guide for trading compute for GPU memory with activation recompute on Megatron Bridge.
nemo-mbridge-perf-cpu-offloadingOpenClaw
NVIDIA guide for validating and applying CPU offloading on Megatron Bridge.
nemo-mbridge-perf-cuda-graphsOpenClaw
NVIDIA guide for validating and applying CUDA graphs on Megatron Bridge.
nemo-mbridge-perf-expert-parallel-overlapOpenClaw
NVIDIA guide for overlapping MoE expert-parallel communication on Megatron Bridge.
nemo-mbridge-perf-hierarchical-context-parallelOpenClaw
NVIDIA operational guide for turning on hierarchical context parallelism for Megatron Bridge.
nemo-mbridge-perf-megatron-fsdpOpenClaw
NVIDIA operational guide for turning on Megatron FSDP for Megatron Bridge.
nemo-mbridge-perf-memory-tuningOpenClaw
NVIDIA guide for lowering peak GPU memory during Megatron Bridge training.
nemo-mbridge-perf-moe-comm-overlapOpenClaw
NVIDIA guide for hiding MoE expert communication behind compute on Megatron Bridge.
nemo-mbridge-perf-moe-dispatcher-selectionOpenClaw
NVIDIA guide for choosing an MoE token dispatcher on Megatron Bridge.
nemo-mbridge-perf-moe-hardware-configsOpenClaw
NVIDIA reference of dated MoE training playbooks sorted by GPU hardware and model.
nemo-mbridge-perf-moe-long-contextOpenClaw
NVIDIA guide for long-context MoE training on Megatron Bridge.
nemo-mbridge-perf-moe-optimization-workflowOpenClaw
NVIDIA workflow for optimizing MoE performance on Megatron Bridge, gated by evidence.
nemo-mbridge-perf-moe-vlm-trainingOpenClaw
NVIDIA guide for training MoE vision-language models on Megatron Bridge.
nemo-mbridge-perf-parallelism-strategiesOpenClaw
NVIDIA operational guide for selecting and mixing parallelism strategies on Megatron Bridge.
nemo-mbridge-perf-sequence-packingOpenClaw
NVIDIA guide for sequence packing and long-context runs on Megatron Bridge.
nemo-mbridge-perf-tp-dp-comm-overlapOpenClaw
NVIDIA operational guide for overlapping tensor, data, and pipeline parallel communication.
nemo-mbridge-recipe-recommenderOpenClaw
NVIDIA skill for recommending and customizing Megatron Bridge recipes for a given model and hardware.
nemo-mbridge-resiliencyOpenClaw
NVIDIA guide to Megatron Bridge resiliency: fault tolerance, stragglers, restarts, and preemption.
nemo-relay-debug-runtime-integrationOpenClaw
NVIDIA skill for debugging missing or wrong NeMo Relay runtime behavior after installation.
nemo-relay-get-startedOpenClaw
NVIDIA skill for newcomers trying NeMo Relay with the simplest supported quick start.
nemo-relay-installOpenClaw
NVIDIA skill for choosing and running the right NeMo Relay install.
nemo-relay-instrument-callsOpenClaw
NVIDIA skill for wrapping tool and LLM calls in NeMo Relay scopes.
nemo-relay-instrument-context-isolationOpenClaw
NVIDIA skill for keeping NeMo Relay scope stacks separate across concurrent work.
nemo-relay-instrument-typed-wrappersOpenClaw
NVIDIA skill for adding typed wrappers, domain types, and provider codecs to NeMo Relay.
nemo-relay-migrate-from-flowOpenClaw
NVIDIA skill for migrating code and docs from NeMo Flow to Relay across languages.
nemo-relay-plugin-adaptive-tuningOpenClaw
NVIDIA skill for configuring and evaluating adaptive plugin behavior in NeMo Relay.
nemo-relay-plugin-buildOpenClaw
NVIDIA skill for building and packaging reusable runtime plugins for NeMo Relay.
nemo-relay-plugin-observabilityOpenClaw
NVIDIA skill for choosing and configuring observability exports in recent NeMo Relay versions.
nemo-retrieverOpenClaw
NVIDIA skill for searching, ingesting, and querying document collections with the NeMo Retriever CLI.
nemo-retriever-mcpOpenClaw
NVIDIA skill for searching and adding documents through the NeMo Retriever MCP server.
nemo-rl-auto-researchOpenClaw
NVIDIA workflow for autonomous NeMo-RL research, from hypothesis testing to open-ended discovery.
nemo-rl-brev-etiquetteOpenClaw
NVIDIA skill for NeMo-RL agents on Brev instances with a small workspace disk.
nemo-rl-docsOpenClaw
NVIDIA skill for NeMo-RL documentation conventions: index updates and docstring format.
nemo-rl-session-memoryOpenClaw
NVIDIA skill for persistent session memory that coding agents can recover.
nemoclaw-user-guideOpenClaw
NVIDIA skill that sends users' AI agents to NemoClaw's docs server and Markdown docs.
nemotron-asr-finetuneOpenClaw
NVIDIA orchestration skill for adapting Nemotron Speech and NeMo ASR to a domain or language.
nemotron-customizeOpenClaw
NVIDIA skill for planning and chaining Nemotron customization steps into pipelines.
nemotron-policy-generatorOpenClaw
NVIDIA skill for generating custom safety policies for Nemotron content-safety guardrails.
nemotron-retrieval-recipesOpenClaw
NVIDIA skill for planning, tuning, and deploying public Nemotron embedding and rerank recipes.
nemotron-speechOpenClaw
NVIDIA skill for routing Nemotron Speech NIM tasks across ASR, TTS, and translation.
nemotron-voice-agent-builderOpenClaw
NVIDIA skill for creating, refining, or fixing real-time voice agents with Pipecat or LiveKit.
nv-generate-ct-rflowOpenClaw
NVIDIA skill for generating synthetic CT scans with masks using the rflow-ct model.
nv-generate-mrOpenClaw
NVIDIA skill for generating synthetic whole-body MRI scans with the rflow-mr model.
nv-generate-mr-brainOpenClaw
NVIDIA skill for generating synthetic brain MRI volumes in several contrasts.
nv-generate-mr-brain-finetuneOpenClaw
NVIDIA skill for fine-tuning the MR-Brain generator on brain MRI NIfTI data.
nv-generate-vae-finetuneOpenClaw
NVIDIA skill for fine-tuning the NV-Generate MAISI VAE on CT or MRI NIfTI datalists.
nv-reason-ctOpenClaw
NVIDIA skill for CT reasoning inference on 3D chest or abdomen scans supplied by users.
nv-reason-cxrOpenClaw
NVIDIA skill for smoke-testing chest X-ray reasoning, by command shape or live run.
nv-segment-ctOpenClaw
NVIDIA skill for CT segmentation with VISTA3D, saving label maps as evidence.
nv-segment-ct-finetuneOpenClaw
NVIDIA skill for softmax fine-tuning of VISTA3D on CT image and label datasets.
nv-segment-ctmrOpenClaw
NVIDIA skill for CT or MRI segmentation that saves label maps as evidence.
nvflare-autoflOpenClaw
NVIDIA skill for agent-driven Auto-FL tuning of an NVFLARE job.
nvflare-autofl-reportOpenClaw
NVIDIA skill for writing the final report for a halted Auto-FL campaign.
nvflare-convert-huggingfaceOpenClaw
NVIDIA skill to convert Hugging Face Trainer or TRL code to a federated NVFLARE job.
nvflare-convert-lightningOpenClaw
NVIDIA skill for converting Lightning training code to a federated NVFLARE job.
nvflare-convert-pytorchOpenClaw
NVIDIA skill for turning plain PyTorch training loops into a federated NVFLARE job.
nvflare-diagnose-jobOpenClaw
NVIDIA skill for finding out why an NVFLARE job failed, hung, or lost clients.
nvflare-fed-statsOpenClaw
NVIDIA skill for computing federated statistics over tabular and image data across NVFLARE sites.
nvflare-orientOpenClaw
NVIDIA skill that routes broad NVFLARE questions and conversions with unclear framework ownership.
nvflare-sharedOpenClaw
NVIDIA internal references and templates that other NVFLARE conversion skills share.
NVIDIA NeMo Fabric Adapter BuilderOpenClaw
NVIDIA skill for building typed third-party adapters for NeMo Fabric.
NVIDIA Skill FinderOpenClaw
NVIDIA skill for finding relevant NVIDIA skills and installing them with approval.
nvmolkit-usageOpenClaw
NVIDIA skill for writing and debugging Python code that uses nvMolKit for GPU cheminformatics.
Omniverse Realtime ViewerOpenClaw
NVIDIA skill for building viewer apps for USD scenes with RTX rendering.
openfold2-nimOpenClaw
NVIDIA skill for predicting monomer protein structures with the OpenFold2 NIM.
openfold3-nimOpenClaw
NVIDIA skill for predicting protein and complex structures with the OpenFold3 NIM.
PAIDF Orchestration — Event Video GenerationOpenClaw
NVIDIA skill for generating event videos and building anomaly datasets.
PAIDF Orchestration — Image Attribute AugmentationOpenClaw
NVIDIA skill for augmenting image attributes and searching images by attribute.
PAIDF Orchestration Environment SetupOpenClaw
NVIDIA skill for auditing, preparing, or deploying the controller and GPU cluster for PAIDF.
paidf-augmentationOpenClaw
NVIDIA skill for PAIDF augmentation configs and remote Cosmos inference jobs.
paidf-auto-labelingOpenClaw
NVIDIA router skill for PAIDF Auto-Labeling planning, cookbooks, prompts, and stage setup.
paidf-cosmos-predictOpenClaw
NVIDIA skill for preparing and running Cosmos Predict video generation on DEFT media.
paidf-curation-and-retrievalOpenClaw
NVIDIA skill for running Cosmos Curator pipelines and nearest-neighbor data mining.
paidf-orchestration-write-dagOpenClaw
NVIDIA skill for writing custom Airflow DAGs for PAIDF Orchestration pipelines.
physical-ai-defect-image-generationOpenClaw
NVIDIA skill for generating defect images with Cosmos AnomalyGen on OSMO.
physical-ai-infrastructure-setup-and-resilient-scalingOpenClaw
NVIDIA skill for setting up, scaling, and hardening physical AI infrastructure on MicroK8s or AKS.
physical-ai-neural-reconstructionOpenClaw
NVIDIA router skill for NuRec and NRE neural reconstruction: USDZ, NCore, 3DGS, and sensor sim.
physical-ai-video-data-augmentationOpenClaw
NVIDIA skill for running video augmentation and auto-labeling on OSMO.
physicsnemo-discoverOpenClaw
NVIDIA official guide for finding the right PhysicsNeMo component for a scientific ML task.
physicsnemo-shard-tensorOpenClaw
NVIDIA official guide for domain parallelism with PhysicsNeMo ShardTensor.
portfolio-optimizationOpenClaw
NVIDIA skill for building, optimizing, and backtesting stock portfolios with cuOpt.
proteinmpnn-nimOpenClaw
NVIDIA skill for designing protein sequences that fit a target backbone using ProteinMPNN.
rag-blueprintOpenClaw
NVIDIA skill for deploying, configuring, troubleshooting, and managing the RAG Blueprint.
rag-evalOpenClaw
NVIDIA skill for filesystem RAG benchmarks scored with RAGAS quality metrics.
rag-perfOpenClaw
NVIDIA skill for benchmarking a deployed RAG Blueprint server with profiling and load tests.
rfdiffusion-nimOpenClaw
NVIDIA skill for designing protein backbones with RFDiffusion through NIM.
RTVI VSS CV ServiceOpenClaw
NVIDIA skill for scaffolding a standalone CV microservice that streams detections into Kafka.
rtvi-cv-customize-modelOpenClaw
NVIDIA guide for replacing the detector used in VSS Alerts 2d_cv verification.
rtvi-vlm-customize-modelOpenClaw
NVIDIA guide for replacing the vision-language model behind VSS Alerts.
rtx-remix-moddingOpenClaw
NVIDIA skill for modding or remastering games through the RTX Remix Toolkit.
skill-card-generatorOpenClaw
NVIDIA skill for writing or refreshing a governance card for an existing agent skill.
tao-analyze-changenet-rcaOpenClaw
NVIDIA TAO skill for deep root-cause investigation of Visual ChangeNet classification runs.
tao-analyze-detection-kpiOpenClaw
NVIDIA TAO skill for object detection KPI analysis against ground truth.
tao-analyze-gaps-od-mapOpenClaw
NVIDIA TAO skill for finding weak images in object detection results.
tao-analyze-gaps-visual-changenetOpenClaw
NVIDIA TAO skill for gap analysis on Visual ChangeNet classify experiments, ranking weak samples.
tao-analyze-gaps-vlm-bcqOpenClaw
NVIDIA TAO skill for finding VLM failures on yes/no classification questions.
tao-artifactsOpenClaw
NVIDIA TAO reference for the JSON schemas behind its execution pipeline without an SDK.
tao-convert-dataset-formatOpenClaw
NVIDIA TAO skill for switching a DAFT dataset to another supported format.
tao-data-ioOpenClaw
NVIDIA TAO skill for staging inputs, credentials, and outputs of TAO jobs.
tao-finetune-clipOpenClaw
NVIDIA TAO skill for CLIP: retrieval, classification, embeddings, and TensorRT deployment.
tao-finetune-cosmos-embedOpenClaw
NVIDIA TAO skill for Cosmos-Embed1 video embeddings, search, and fine-tuning.
tao-finetune-cosmos-reasonOpenClaw
NVIDIA TAO skill for the shared Cosmos3 frontend for Cosmos-RL and framework training.
tao-finetune-huggingface-modelOpenClaw
NVIDIA TAO skill for fine-tuning a Hugging Face model when no TAO model skill fits.
tao-finetune-nv-tesseract-ad-diffusionOpenClaw
NVIDIA TAO skill for diffusion-based time series anomaly detection with NV-Tesseract.
tao-finetune-nv-tesseract-forecastingOpenClaw
NVIDIA TAO skill for transformer-based time series forecasting with NV-Tesseract.
tao-finetune-video-clipOpenClaw
NVIDIA TAO skill for InternVideo2-CLIP video retrieval, LoRA tuning, and deployment.
tao-generate-anomaliesOpenClaw
NVIDIA TAO skill for PAIDF AnomalyGen: train, generate, score, and tune synthetic anomalies.
tao-generate-image-embeddingsOpenClaw
NVIDIA TAO skill for converting a parquet of image paths into embeddings.
tao-generate-image-groundingOpenClaw
NVIDIA TAO skill for a two-step pipeline that grounds referring expressions to bounding boxes.
tao-generate-referring-expressionsOpenClaw
NVIDIA TAO skill for creating referring expressions from images and KITTI boxes.
tao-generate-video-reasoning-annotationsOpenClaw
NVIDIA TAO skill for producing reasoning training data from raw video.
tao-launch-workflowOpenClaw
NVIDIA TAO gate that must run before launching any TAO workflow on a platform.
tao-list-capabilitiesOpenClaw
NVIDIA TAO skill for describing the plugin's abilities from its packaged manifests.
tao-mine-aoi-imagesOpenClaw
NVIDIA TAO skill for embedding and mining source images for VCN AOI iterations.
tao-mine-nearest-neighborsOpenClaw
NVIDIA TAO skill for finding source samples closest to targets using embeddings.
tao-mine-od-imagesOpenClaw
NVIDIA TAO skill for one-to-one source image mining in detection workflows.
tao-port-huggingface-modelOpenClaw
NVIDIA TAO skill for bringing a Hugging Face vision model into TAO Toolkit.
tao-route-visual-changenet-samplesOpenClaw
NVIDIA TAO skill for sorting weak VCN samples into subsets per augmentation module.
tao-run-automlOpenClaw
NVIDIA TAO skill for AutoML and hyperparameter search on TAO networks in containers.
tao-run-automl-deft-pipelineOpenClaw
NVIDIA TAO skill for the three-phase AOI training pipeline: AutoML, DEFT loop, AutoML refinement.
tao-run-deft-aoiOpenClaw
NVIDIA TAO skill for the end-to-end DEFT improvement loop on PCB inspection models.
tao-run-deft-aoi-cosmos3OpenClaw
NVIDIA TAO skill for the disk-based DEFT AOI loop on Cosmos Reason 3.
tao-run-deft-cr-its-miningOpenClaw
NVIDIA TAO skill for mining-driven DEFT improvement on binary ITS video questions.
tao-run-deft-object-detectionOpenClaw
NVIDIA TAO skill for repeated DEFT augmentation loops on Grounding DINO detectors.
tao-run-deft-pasOpenClaw
NVIDIA TAO skill for iteratively improving CLIP or SigLIP retrieval on attribute data.
tao-run-inference-serviceOpenClaw
NVIDIA TAO skill for starting, querying, and stopping a network-specific inference microservice.
tao-run-on-brevOpenClaw
NVIDIA TAO skill for running TAO containers on Brev GPU instances.
tao-run-on-dockerOpenClaw
NVIDIA TAO skill for running TAO jobs through Docker, local or remote.
tao-run-on-kubernetesOpenClaw
NVIDIA TAO skill for running TAO container jobs on Kubernetes with GPU scheduling.
tao-run-on-slurmOpenClaw
NVIDIA TAO skill for running TAO jobs on a SLURM GPU cluster through SSH.
tao-run-on-virtualenvOpenClaw
NVIDIA TAO skill for running training or eval scripts in a local virtualenv without containers.
tao-setupOpenClaw
NVIDIA TAO skill for first-run setup and the map of the TAO skill bank.
tao-setup-nvidia-gpu-hostOpenClaw
NVIDIA TAO skill for verifying and installing GPU driver, CUDA, and toolkit versions.
tao-train-action-recognitionOpenClaw
NVIDIA TAO skill for training video action recognition on RGB, optical flow, or joint inputs.
tao-train-bevfusionOpenClaw
NVIDIA TAO skill for training BEVFusion 3D detection from LiDAR and camera data.
tao-train-centerposeOpenClaw
NVIDIA TAO skill for training CenterPose keypoint and object pose estimation.
tao-train-codetrOpenClaw
NVIDIA TAO skill for training Co-DETR (CoDINO) object detectors.
tao-train-deformable-detrOpenClaw
NVIDIA TAO skill for training Deformable DETR 2D object detectors.
tao-train-depth-anything-v2OpenClaw
NVIDIA TAO skill for training monocular depth estimation with Depth Anything v2.
tao-train-dinoOpenClaw
NVIDIA TAO skill for training DINO transformer detectors with denoising anchor boxes.
tao-train-dinov3OpenClaw
NVIDIA TAO skill for continual self-supervised pretraining of DINOv3 backbones on unlabeled images.
tao-train-fast-foundation-stereoOpenClaw
NVIDIA TAO skill for training FastFoundationStereo real-time stereo depth models.
tao-train-foundation-stereoOpenClaw
NVIDIA TAO skill for training FoundationStereo disparity models for 3D reconstruction.
tao-train-grounding-dinoOpenClaw
NVIDIA TAO skill for training Grounding DINO text-prompted open-set detectors.
tao-train-image-classificationOpenClaw
NVIDIA TAO skill for training PyTorch image classifiers with distillation and quantization.
tao-train-mask-auto-encoderOpenClaw
NVIDIA TAO skill for MAE pretraining and fine-tuning on unlabeled images.
tao-train-mask-auto-labelOpenClaw
NVIDIA TAO skill for weakly supervised segmentation masks from points or boxes.
tao-train-mask-grounding-dinoOpenClaw
NVIDIA TAO skill for training Mask Grounding DINO text-guided instance segmentation.
tao-train-mask2formerOpenClaw
NVIDIA TAO skill for training Mask2Former panoptic, instance, and semantic segmentation.
tao-train-metric-learning-recognitionOpenClaw
NVIDIA TAO skill for training embeddings for fine-grained recognition and matching.
tao-train-nvdinov2OpenClaw
NVIDIA TAO skill for training NVDINOv2 self-supervised visual features without labels.
tao-train-nvpanoptix3dOpenClaw
NVIDIA TAO skill for training NVPanoptix3D 3D scene models from posed RGB images.
tao-train-ocdnetOpenClaw
NVIDIA TAO skill for training OCDNet scene text detectors.
tao-train-ocrnetOpenClaw
NVIDIA TAO skill for training OCRNet scene text recognizers.
tao-train-oneformerOpenClaw
NVIDIA TAO skill for training OneFormer universal segmentation with task-conditioned queries.
tao-train-optical-inspectionOpenClaw
NVIDIA TAO skill for training Siamese-network defect detection on image pairs.
tao-train-pointpillarsOpenClaw
NVIDIA TAO skill for training PointPillars LiDAR 3D detectors.
tao-train-pose-classificationOpenClaw
NVIDIA TAO skill for training ST-GCN classifiers on skeleton pose sequences.
tao-train-reidOpenClaw
NVIDIA TAO skill for training person re-identification embeddings across camera views.
tao-train-rtdetrOpenClaw
NVIDIA TAO skill for training RT-DETR real-time object detectors.
tao-train-segformerOpenClaw
NVIDIA TAO skill for training SegFormer semantic segmentation models.
tao-train-single-stepOpenClaw
NVIDIA TAO skill for a plain train, evaluate, and export run on any TAO model.
tao-train-sparse4dOpenClaw
NVIDIA TAO skill for training Sparse4D 3D detection and tracking from multiple cameras.
tao-train-visual-changenetOpenClaw
NVIDIA TAO skill for Visual ChangeNet pass/fail classification and change segmentation in AOI.
tao-validate-dataset-formatOpenClaw
NVIDIA TAO skill for checking DAFT datasets for layout, schema, and reference problems.
tao-validate-recipe-transferOpenClaw
NVIDIA TAO skill for moving a vision paper's code and recipe onto a customer dataset.
tilegym-adding-cutile-kernelOpenClaw
NVIDIA skill for adding a cuTile kernel operator to TileGym with tests and benchmark.
tilegym-converting-cutile-to-juliaOpenClaw
NVIDIA skill for converting cuTile Python kernels into cuTile.jl Julia versions.
tilegym-converting-cutile-to-tritonOpenClaw
NVIDIA skill for converting cuTile GPU kernels to Triton, including debugging.
tilegym-cutile-autotuningOpenClaw
NVIDIA skill for adding, changing, or debugging autotuning in cuTile kernels.
tilegym-cutile-pythonOpenClaw
NVIDIA skill for writing high-performance GPU kernels with cuTile's tile-based model.
tilegym-improve-cutile-kernel-perfOpenClaw
NVIDIA skill for iteratively speeding up cuTile kernels via profiling and tuning.
tilegym-monkey-patch-kernels-to-transformersOpenClaw
NVIDIA skill for wiring TileGym kernels into transformers model classes.
Train Workflow Policy with RLOpenClaw
NVIDIA skill for training, evaluating, and exporting reinforcement learning policies in a Workflow.
USD Performance TuningOpenClaw
NVIDIA skill for diagnosing and optimizing OpenUSD scene performance.
vss-ask-videoOpenClaw
NVIDIA skill for putting a new visual question about a clip to the VSS agent.
vss-deploy-dense-captioningOpenClaw
NVIDIA skill for deploying the standalone RT-VLM captioner and calling its REST API.
vss-deploy-detection-tracking-2dOpenClaw
NVIDIA skill for deploying and operating the 2D detection and tracking microservice for VSS.
vss-deploy-detection-tracking-3dOpenClaw
NVIDIA skill for deploying MV3DT 3D detection with BEV fusion across calibrated cameras.
vss-deploy-profileOpenClaw
NVIDIA skill for selecting, configuring, deploying, and debugging a VSS profile.
vss-deploy-video-embeddingOpenClaw
NVIDIA skill for deploying and integrating the VSS video embedding microservice.
vss-generate-video-calibrationOpenClaw
NVIDIA skill for running AutoMagicCalib camera calibration on videos or the sample data.
vss-generate-video-reportOpenClaw
NVIDIA skill for producing VSS analysis reports from per-clip VLM or incident ranges.
vss-manage-alertsOpenClaw
NVIDIA skill for VSS alert workflows: monitoring, Slack notices, incident queries, camera onboarding.
vss-manage-video-io-storageOpenClaw
NVIDIA skill for calling the VIOS REST API for sensors, timelines, clips, and snapshots.
vss-query-analyticsOpenClaw
NVIDIA skill for reading video analytics data through the VSS VA-MCP server.
vss-search-archiveOpenClaw
NVIDIA skill for fusion search over archived video and ingesting files or streams.
vss-setup-behavior-analyticsOpenClaw
NVIDIA skill for deploying the behavior analytics service by itself.
vss-setup-video-analytics-apiOpenClaw
NVIDIA skill for deploying the VSS video analytics REST service by itself.
vss-summarize-videoOpenClaw
NVIDIA skill for summarizing recorded video with the LVS microservice and VLM fallback.
warp-compile-time-optimizerOpenClaw
NVIDIA skill for cutting compile and startup time in Warp code.
warp-debug-gradientsOpenClaw
NVIDIA skill for fixing incorrect gradients in differentiable simulation code.
warp-evalOpenClaw
NVIDIA skill for judging whether existing slow code is a good Warp candidate.