nvidia-tao/tao-skill-bank

nvidia-tao/tao-skill-bank · AGENTS.md

nvidia-tao/tao-skill-bank · integrations/nemoclaw/AGENTS.md

nvidia-tao/tao-skill-bank · skills/applications/tao-analyze-changenet-rca/SKILL.md · Performs deep Root Cause Analysis (RCA) on NVIDIA TAO Visual ChangeNet classification experiments with

nvidia-tao/tao-skill-bank · skills/applications/tao-finetune-huggingface-model/SKILL.md · Fine-tune any HuggingFace CV / VLM / LLM model on local NVIDIA GPUs inside an NGC PyTorch container when no dedicated…

nvidia-tao/tao-skill-bank · skills/applications/tao-port-huggingface-model/SKILL.md · Integrate a HuggingFace Computer Vision model into the NVIDIA TAO Toolkit ecosystem (tao-core config, tao-pytorch…

nvidia-tao/tao-skill-bank · skills/applications/tao-run-automl-deft-pipeline/SKILL.md · Run the canonical NVIDIA AOI three-phase training pipeline — Phase 1 AutoML baseline (HPO), Phase 2 DEFT loop (RCA →…

nvidia-tao/tao-skill-bank · skills/applications/tao-run-automl/SKILL.md · Run container-backed AutoML / hyperparameter optimization (HPO) for NVIDIA TAO networks using AutoMLRunner. Handles…

nvidia-tao/tao-skill-bank · skills/applications/tao-run-deft-aoi-cosmos3/SKILL.md · Run the disk-backed DEFT AOI improvement loop for NVIDIA Cosmos Reason 3 / Cosmos3 models, using Nano by default and…

nvidia-tao/tao-skill-bank · skills/applications/tao-run-deft-aoi/SKILL.md · Run the full DEFT AOI improvement loop for NVIDIA TAO VisualChangeNet / ChangeNet PCB inspection models: baseline…

nvidia-tao/tao-skill-bank · skills/applications/tao-run-deft-cr-its-mining/SKILL.md · Run the mining-based DEFT improvement workflow for ITS Cosmos-Reason binary video questions, focused on the…

nvidia-tao/tao-skill-bank · skills/applications/tao-run-deft-object-detection/SKILL.md · Run the full DEFT smart-data-augmentation loop for NVIDIA TAO Grounding DINO object detection: zero-shot baseline…

nvidia-tao/tao-skill-bank · skills/applications/tao-run-deft-od-aoi/SKILL.md · Run a binary industrial-inspection DEFT loop with TAO RT-DETR: measure on frozen KPI/test roles, mine gap-similar real…

nvidia-tao/tao-skill-bank · skills/applications/tao-run-deft-pas/SKILL.md · Run iterative improvement for NVIDIA TAO CLIP / SigLIP image-text retrieval on attribute-labelled data. Use when a…

nvidia-tao/tao-skill-bank · skills/applications/tao-run-dinov3-ssl-deft/SKILL.md · Run iterative DINOv3 SSL DEFT data selection and retraining from an unlabeled pool, using GRIT or multi-task weakness…

nvidia-tao/tao-skill-bank · skills/applications/tao-run-inference-service/SKILL.md · Start, query, and stop a network-specific TAO inference microservice ({network_arch}-inference-microservice) by…

nvidia-tao/tao-skill-bank · skills/applications/tao-train-single-step/SKILL.md · Standard single-step train/eval/export workflow for any TAO model. Use when training a TAO model on a dataset

nvidia-tao/tao-skill-bank · skills/applications/tao-validate-recipe-transfer/SKILL.md · Port a published computer vision paper's official code and training recipe onto a customer's own dataset, or diagnose…

nvidia-tao/tao-skill-bank · skills/core/tao-artifacts/SKILL.md · The contract home for TAO's SDK-free execution pipeline — authoritative JSON Schemas for the four typed artifacts…

nvidia-tao/tao-skill-bank · skills/core/tao-launch-workflow/SKILL.md · The mandatory pre-launch gate and four-verb execution contract for every TAO workflow or action. Invoke BEFORE…

nvidia-tao/tao-skill-bank · skills/core/tao-list-capabilities/SKILL.md · Answer what the TAO Skill Bank plugin can do by generating the response from packaged application, data, model, AutoML…

nvidia-tao/tao-skill-bank · skills/core/tao-setup/SKILL.md · One-time session setup and orchestration map for the TAO skill bank. Run this first when the TAO skills were installed…

nvidia-tao/tao-skill-bank · skills/core/tao-token-efficient-execution/SKILL.md · Token-efficient execution kit for repeated TAO agentic workflows. Compiles a skill into stage cards once (with a strong…

nvidia-tao/tao-skill-bank · skills/data/paidf-cosmos-predict/SKILL.md · Prepare and run PAIDF Cosmos Predict video generation for DEFT media samples.

nvidia-tao/tao-skill-bank · skills/data/tao-analyze-detection-kpi/SKILL.md · Run TAO Data Services KPI analysis for object detection, comparing inference annotations against ground truth to…

nvidia-tao/tao-skill-bank · skills/data/tao-analyze-gaps-od-map/SKILL.md · Run TAO Data Services object-detection gap analysis from ground-truth and inference annotations. Use when an object…

nvidia-tao/tao-skill-bank · skills/data/tao-analyze-gaps-visual-changenet/SKILL.md · Performs gap analysis on NVIDIA TAO VCN Classify (Visual Component Net) experiments by invoking the pinned TAO…

nvidia-tao/tao-skill-bank · skills/data/tao-analyze-gaps-vlm-bcq/SKILL.md · Extract false-positive and false-negative gaps from VLM binary-classification-question (BCQ, yes/no) predictions.

nvidia-tao/tao-skill-bank · skills/data/tao-convert-dataset-format/SKILL.md · Run `tao-daft convert` to convert NVIDIA TAO DAFT datasets between supported formats. Do not use for non-DAFT data.

nvidia-tao/tao-skill-bank · skills/data/tao-finetune-anomalygennext/SKILL.md · Validate a user AnomalyGenNext dataset and recipe, freeze its validation set, and prepare a canonical texture…

nvidia-tao/tao-skill-bank · skills/data/tao-generate-anomalies/SKILL.md · Full PAIDF AnomalyGen pipeline — fine-tune on a new anomaly dataset, generate synthetic anomaly images (SDG), evaluate…

nvidia-tao/tao-skill-bank · skills/data/tao-generate-image-embeddings/SKILL.md · Run TAO Data Services image embedding to turn a parquet of image filepaths into an embedding parquet using CLIP…

nvidia-tao/tao-skill-bank · skills/data/tao-generate-image-grounding/SKILL.md · "Two-step image grounding pipeline: extracts referring expressions from (image, caption) pairs and grounds them

nvidia-tao/tao-skill-bank · skills/data/tao-generate-od-defects/SKILL.md · Run AnomalyGenNext 1.1 inference from a native testcase or a prepared generation plan, then publish validated…

nvidia-tao/tao-skill-bank · skills/data/tao-generate-referring-expressions/SKILL.md · "Four-step image referring-expression pipeline: turns images plus KITTI bounding-box labels into region

nvidia-tao/tao-skill-bank · skills/data/tao-generate-video-reasoning-annotations/SKILL.md · Multi-step video annotation pipeline that turns raw videos into Chain-of-Thought training data — multi-level captions…

nvidia-tao/tao-skill-bank · skills/data/tao-mine-aoi-images/SKILL.md · Runs the DEFT embed-then-mine workflow for VCN AOI iterations — embeds the gap-analysis target parquet, embeds a source…

nvidia-tao/tao-skill-bank · skills/data/tao-mine-nearest-neighbors/SKILL.md · Run TAO Data Services TMM nearest-neighbor mining from embedding parquet files. Use when a workflow needs to mine…

nvidia-tao/tao-skill-bank · skills/data/tao-mine-od-images/SKILL.md · Run TAO Data Services TMM unique-neighbor matching mining from embedding parquet files for object detection workflows…

nvidia-tao/tao-skill-bank · skills/data/tao-prepare-anomalygennext-inputs/SKILL.md · Prepare normalized object-detection false negatives, source masks, and clean-image embedding inputs for AnomalyGenNext…

nvidia-tao/tao-skill-bank · skills/data/tao-route-visual-changenet-samples/SKILL.md · Routes the weakest VCN samples (output of `tao-analyze-gaps-visual-changenet`) into per-augmentation-module

nvidia-tao/tao-skill-bank · skills/data/tao-validate-dataset-format/SKILL.md · Run `tao-daft validate` to check NVIDIA TAO DAFT datasets for structure, schema, and cross-reference errors. Do

nvidia-tao/tao-skill-bank · skills/models/tao-finetune-clip/SKILL.md · CLIP vision-language model for image-text retrieval, zero-shot classification, embedding extraction, ONNX

nvidia-tao/tao-skill-bank · skills/models/tao-finetune-cosmos-embed/SKILL.md · Cosmos-Embed1 video-text embedding for text-to-video retrieval, video-to-video search, semantic deduplication, and…

nvidia-tao/tao-skill-bank · skills/models/tao-finetune-cosmos-reason/SKILL.md · Shared Cosmos3 frontend that explicitly routes Cosmos Framework and Cosmos-RL, validates runtime…

nvidia-tao/tao-skill-bank · skills/models/tao-finetune-kumo-anomaly/SKILL.md · Kumo-Anomaly — diffusion-based anomaly detection and fine-tuning for multivariate time series. Use when the user asks…

nvidia-tao/tao-skill-bank · skills/models/tao-finetune-kumo-forecast/SKILL.md · Kumo-Forecast — transformer-based multivariate time series forecasting with DARR (context-enhanced kNN retrieval)…

nvidia-tao/tao-skill-bank · skills/models/tao-finetune-video-clip/SKILL.md · InternVideo2-CLIP L14 (TAO video_clip) for video-text retrieval, zero-shot classification, embedding extraction, LoRA…

nvidia-tao/tao-skill-bank · skills/models/tao-train-action-recognition/SKILL.md · Action recognition from video sequences. Supports RGB, optical flow, and joint (multi-stream) input types for

nvidia-tao/tao-skill-bank · skills/models/tao-train-bevfusion/SKILL.md · BEVFusion for multi-sensor 3D object detection. Fuses LiDAR point clouds and camera images in bird's-eye-view

nvidia-tao/tao-skill-bank · skills/models/tao-train-centerpose/SKILL.md · CenterPose for keypoint / pose estimation. Detects object centers and regresses keypoint locations for 6-DoF

nvidia-tao/tao-skill-bank · skills/models/tao-train-codetr/SKILL.md · Co-DETR (CoDINO) for object detection. A DETR-family detector with collaborative hybrid

nvidia-tao/tao-skill-bank · skills/models/tao-train-deformable-detr/SKILL.md · Deformable DETR for 2D object detection. Uses deformable attention for efficient multi-scale feature processing,

nvidia-tao/tao-skill-bank · skills/models/tao-train-depth-anything-v2/SKILL.md · Monocular depth estimation using Metric Depth Anything v2 or Relative Depth Anything architectures. Predicts

nvidia-tao/tao-skill-bank · skills/models/tao-train-dino/SKILL.md · DINO (DETR with Improved DeNoising Anchor Boxes) for 2D object detection. Transformer-based detector with

nvidia-tao/tao-skill-bank · skills/models/tao-train-dinov3/SKILL.md · DINOv3 continual self-supervised pre-training. Domain-adapts public DINOv3 ViT backbones

nvidia-tao/tao-skill-bank · skills/models/tao-train-fast-foundation-stereo/SKILL.md · Real-time stereo depth estimation using FastFoundationStereo (FFS), the distilled bp2 commercial variant of

nvidia-tao/tao-skill-bank · skills/models/tao-train-foundation-stereo/SKILL.md · Stereo depth estimation using FoundationStereo. Predicts disparity maps from stereo image pairs for 3D

nvidia-tao/tao-skill-bank · skills/models/tao-train-grounding-dino/SKILL.md · Grounding DINO for open-set object detection. Combines DINO-style detection with a BERT text encoder for

nvidia-tao/tao-skill-bank · skills/models/tao-train-image-classification/SKILL.md · PyTorch-based TAO image classification. Supports a wide range of backbones (FAN, EfficientNet, ResNet, etc.)

nvidia-tao/tao-skill-bank · skills/models/tao-train-mask-auto-encoder/SKILL.md · Masked Auto-Encoder (MAE) for self-supervised pretraining and fine-tuning. Masks random patches and reconstructs

nvidia-tao/tao-skill-bank · skills/models/tao-train-mask-auto-label/SKILL.md · MAL (Mask Auto-Label) for weakly-supervised segmentation. Produces segmentation masks from minimal annotations

nvidia-tao/tao-skill-bank · skills/models/tao-train-mask-grounding-dino/SKILL.md · Mask Grounding DINO for grounded instance segmentation. Extends Grounding DINO with a mask-prediction head for

nvidia-tao/tao-skill-bank · skills/models/tao-train-mask2former/SKILL.md · Mask2Former for universal image segmentation (panoptic, instance, and semantic). Transformer-based with

nvidia-tao/tao-skill-bank · skills/models/tao-train-metric-learning-recognition/SKILL.md · Metric-learning recognition (ml-recog) for fine-grained visual recognition. Learns embeddings for

nvidia-tao/tao-skill-bank · skills/models/tao-train-nvdinov2/SKILL.md · NVDINOv2 for self-supervised visual representation learning. Trains vision transformers via self-distillation

nvidia-tao/tao-skill-bank · skills/models/tao-train-nvpanoptix3d/SKILL.md · NVPanoptix3D for panoptic 3D scene reconstruction from posed RGB images. Produces 3D panoptic segmentation

nvidia-tao/tao-skill-bank · skills/models/tao-train-ocdnet/SKILL.md · OCDNet for scene text detection. Detects arbitrary-oriented text regions in natural images using a

nvidia-tao/tao-skill-bank · skills/models/tao-train-ocrnet/SKILL.md · OCRNet for scene text recognition. Recognizes text content from cropped text-region images and supports CTC

nvidia-tao/tao-skill-bank · skills/models/tao-train-oneformer/SKILL.md · OneFormer for universal image segmentation. Unifies panoptic, instance, and semantic segmentation with a

nvidia-tao/tao-skill-bank · skills/models/tao-train-optical-inspection/SKILL.md · Optical Inspection for defect detection using Siamese networks. Compares image pairs to detect manufacturing

nvidia-tao/tao-skill-bank · skills/models/tao-train-pointpillars/SKILL.md · PointPillars for 3D object detection from LiDAR point clouds. Encodes point clouds into a pseudo-image via a

nvidia-tao/tao-skill-bank · skills/models/tao-train-pose-classification/SKILL.md · Pose classification using ST-GCN (Spatial Temporal Graph Convolutional Network). Classifies skeleton sequences

nvidia-tao/tao-skill-bank · skills/models/tao-train-reid/SKILL.md · Person re-identification (ReID). Learns discriminative embeddings to match the same person across different

nvidia-tao/tao-skill-bank · skills/models/tao-train-rtdetr/SKILL.md · RT-DETR (Real-Time DEtection TRansformer) for 2D object detection. Designed for real-time inference with

nvidia-tao/tao-skill-bank · skills/models/tao-train-segformer/SKILL.md · SegFormer for semantic segmentation. Lightweight transformer-based architecture with hierarchical feature

nvidia-tao/tao-skill-bank · skills/models/tao-train-sparse4d/SKILL.md · Prepare calibrated multi-camera data and train, evaluate, run inference, export, or quantize TAO Sparse4D 3D detection…

nvidia-tao/tao-skill-bank · skills/models/tao-train-visual-changenet/SKILL.md · Visual ChangeNet for binary image classification and segmentation in AOI defect detection. Use when training,

nvidia-tao/tao-skill-bank · skills/platform/tao-data-io/SKILL.md · The data-mover for TAO jobs — decides the storage tier (A pre-positioned mount with zero fetch / B volume-from-S3 / C…

nvidia-tao/tao-skill-bank · skills/platform/tao-run-on-brev/SKILL.md · Run a TAO training/evaluation/inference container on an NVIDIA Brev GPU instance. Instance provisioning…

nvidia-tao/tao-skill-bank · skills/platform/tao-run-on-docker/SKILL.md · The Docker execution platform for TAO jobs — a local daemon or a remote GPU box via DOCKER_HOST=ssh://user@host…

nvidia-tao/tao-skill-bank · skills/platform/tao-run-on-kubernetes/SKILL.md · Kubernetes execution platform — submits TAO container jobs as k8s Jobs with NVIDIA GPU scheduling; single-pod

nvidia-tao/tao-skill-bank · skills/platform/tao-run-on-slurm/SKILL.md · Remote SLURM GPU cluster execution over SSH with sbatch/srun, Pyxis/Enroot containers, and Lustre-backed

nvidia-tao/tao-skill-bank · skills/platform/tao-run-on-virtualenv/SKILL.md · Run a Python training/eval script directly in an existing local virtualenv — no docker, no container. Implements the…

nvidia-tao/tao-skill-bank · skills/platform/tao-setup-nvidia-gpu-host/SKILL.md · Host setup for TAO GPU backends. Checks and, after user approval, installs minimum-compatible NVIDIA driver, CUDA…

nvidia-tao/tao-skill-bank · templates/skill-skeleton/data/SKILL.md · One-to-three-sentence description of what this data transformation does. Use when the user asks to…

nvidia-tao/tao-skill-bank · templates/skill-skeleton/minimal/SKILL.md · One-to-three-sentence description of what the skill does and when to use it. Use when the user asks to…

nvidia-tao/tao-skill-bank · templates/skill-skeleton/model/SKILL.md · One-to-three-sentence description of what the model does and when to use it. Use when the user asks to "fine-tune…

nvidia-tao/tao-skill-bank · templates/skill-skeleton/platform/SKILL.md · Where and how GPU jobs run on this platform. One-to-three-sentence summary. Use when the user asks to "deploy on…

nvidia-tao/tao-skill-bank · templates/skill-skeleton/workflow/SKILL.md · Top-level workflow orchestrator for <use case>. Runs an end-to-end pipeline of <N> stages: <stage 1> → <stage 2> → … →…