Latest Agent Skills

Create job stories using the 'When [situation], I want to [motivation], so I can [outcome]' format with detailed acceptance criteria, focusing on user situations and outcomes for JTBD-style backlog items.

Run a pre-mortem risk analysis on a PRD or launch plan, categorizing risks as Tigers (real problems), Paper Tigers (overblown concerns), and Elephants (unspoken worries), then classifying by urgency and creating action plans.

Create a comprehensive Product Requirements Document using an 8-section template covering problem, objectives, segments, value propositions, solution, and release planning.

Draft detailed Non-Disclosure Agreements between two parties, covering information types, jurisdiction, and clauses needing legal review.

Facilitate a structured sprint retrospective to surface insights and produce actionable improvements with owners and deadlines.

Create product backlog items in Why-What-Acceptance format — independent, valuable, testable items with strategic context.

Provides Vue 3 Options API best practices, TypeScript integration tips, and solutions to common gotchas.

Create library-grade Vue composables that accept maybe-reactive inputs (MaybeRef/MaybeRefOrGetter) so callers can pass plain values, refs, or getters. Normalize inputs with toValue()/toRef() inside reactive effects for predictable behavior.

Provides best practices, patterns, and solutions for testing Vue.js applications, covering Vitest, Vue Test Utils, component testing, mocking, and Playwright for E2E testing.

Provides Vue Router 4 patterns, navigation guards, route params, and route-component lifecycle interactions.

Learn Pinia best practices, common gotchas, and state management patterns for Vue applications.

Provides guidance on JSX syntax differences between Vue and React, including attribute naming and plugin configuration.

Provides Vue.js best practices and workflow guidance, including architecture planning, component design, reactivity patterns, and performance optimization for Vue 3 projects.

Diagnose and fix Vue 3 runtime errors, warnings, async failures, and SSR/hydration issues with targeted debugging guides.

Validate and configure packed sequences and long-context training in Megatron-Bridge, distinguishing offline packed SFT for LLMs from in-batch packing for VLMs with correct context parallelism constraints.

Provides practical guidance for training Mixture-of-Experts Vision-Language Models in Megatron Bridge, comparing FSDP and 3D-parallel approaches with lessons from recent multimodal experiments.

Provides guidance for training Mixture-of-Experts models with long context windows, covering context parallelism sizing, selective recomputation, dispatcher choices, and practical patterns from recent experiments.

Optimizes MoE expert-parallel communication overlap in Megatron Bridge, covering dispatch/combine overlap, flex dispatcher backends, and expert wgrad scheduling.

Solve LP, MILP, and QP problems using the cuOpt C API with a consistent build pattern and core calls.

Identifies the weakest samples per ground-truth label in NVIDIA TAO VCN Classify experiments by running a Docker container that performs threshold sweep, weakness scoring, and per-lighting expansion, then surfaces top-K weak samples for downstream augmentation or relabeling.

Routes NuRec/Neural Reconstruction requests to the correct upstream NVIDIA skill (datasets, conversion, training, rendering, object harvesting, frame cleanup).

Create and modify NeMo AutoModel training and evaluation recipes, including YAML structure, builders, and execution flow.

Bootstraps a clinical ASR evaluation environment by verifying NVIDIA_API_KEY, installing Python dependencies, and running a smoke test against hosted TTS/ASR services.

Curates clinical-specialty term lists, generates IPA-tagged synthetic audio via TTS, and produces NeMo-format manifests for ASR benchmark evaluation.

Deploy and operate the RTVI-CV-3D microservice for multi-camera 3D detection and tracking, supporting sample datasets, custom videos, and RTSP streams.

Deploy, debug, and operate the RTVI-CV 2D detection/tracking microservice and call its REST API for stream management, health checks, and metrics.

Trains, evaluates, exports, and runs inference for FastFoundationStereo (FFS) stereo depth estimation models, a distilled variant of FoundationStereo with lower latency.

Diagnose and optimize slow-loading, high-memory, or low-FPS USD scenes using a structured workflow with profiling, validation, and mutation phases.

Configure and validate CPU offloading for Megatron Bridge training, including activation offloading and optimizer state offloading with HybridDeviceOptimizer.

Run Megatron-LM (MLM) and Megatron Bridge training with mock or real data, covering correlation testing, available recipes, and multi-GPU examples.

Score a clinical ASR manifest against a chosen NIM, produce a five-section KER leaderboard, and route the user via a post-eval decision tree.

Indexes Megatron Bridge recipes and recommends the best starting config based on model, GPU count, and training goal.

Reduces peak GPU memory in Megatron Bridge training by applying expandable segments, parallelism resizing, activation recompute, and CPU offloading constraints.

Enables Megatron Fully Sharded Data Parallel in Megatron-Bridge with configuration overrides, code anchors, pitfalls, and verification steps.

Guides implementation of new model architectures in NeMo AutoModel through five phases: discovery, implementation, registration, validation, and testing.

Assesses NumPy code for cuPyNumeric migration readiness by analyzing source code against an API support manifest and GPU-scaling idioms, producing a structured verdict with per-finding reasoning.

Runs AutoMagicCalib calibration on local MP4s, RTSP streams, or a bundled sample dataset, and deploys the AMC microservice when needed.

Train, evaluate, distill, quantize, export, and run inference for PyTorch-based TAO image classification models with support for multiple backbones.

Fine-tune HuggingFace CV, VLM, or LLM models on local NVIDIA GPUs using an NGC PyTorch container, with support for full or LoRA training, dataset handling, and optional model push to the Hub.

Configure NeMo AutoModel job launches for interactive runs, Slurm clusters, and SkyPilot cloud execution.

Import object detection models from HuggingFace or NVIDIA NGC into a DeepStream pipeline with automated ONNX download, TensorRT engine build, custom parser, multi-stream benchmark, and PDF report generation.

Load sharded datasets (npy, Parquet, HDF5, raw binary) into distributed cuPyNumeric arrays using manual partitioning and Legate task launches.

Deploys the vss-video-analytics-api REST service standalone with config, data-log bind, and optional Elasticsearch/Kafka connectivity.

Train, evaluate, export, and run inference for pose classification models using ST-GCN on skeleton keypoint sequences.

Trains, evaluates, exports, quantizes, and runs inference for a Mask Grounding DINO model for open-set instance segmentation guided by text prompts.

Finetunes the NV-Generate-CTMR MR-brain diffusion UNet from user-supplied NIfTI training volumes using a wrapper that stages configs and delegates to upstream scripts.

Validate and use CUDA graph capture in Megatron Bridge, including local full-iteration graphs and Transformer Engine scoped graphs for attention, MLP, and MoE modules.

Convert single-node PyTorch distributed scripts into multi-node Slurm sbatch jobs and debug common multi-node failures, covering srun-native and torch.distributed approaches, container setup, NCCL timeouts, and interactive allocation.

Deploy the behavior-analytics service standalone with a chosen entrypoint, config source, and optional calibration, without the full warehouse stack.

Add a new cuTile GPU kernel operator to TileGym, covering dispatch registration, backend implementation, exports, tests, and benchmarks.

Trains, evaluates, exports, and runs inference for Siamese-network-based optical inspection models to detect manufacturing defects and quality issues in image pairs.

Train, evaluate, export, and run inference on TAO action-recognition models for classifying temporal actions in video clips using RGB, optical flow, or joint input.

Runs a three-phase AOI training pipeline: AutoML HPO baseline, DEFT iterative data improvement, and AutoML refinement on the augmented dataset.

Generates phrase-grounded bounding box annotations from image-caption pairs using a VLM, producing cleaned captions, referring expressions, and pixel-space bounding boxes.

Lint and format Python code for Megatron-LM using ruff, black, isort, pylint, and mypy, with commands for autoformatting and import ordering.

Manage VIOS and NvStreamer REST API operations for video input/output and storage, including sensors, streams, snapshots, clips, and recordings.

Deploy a standalone RT-VLM dense-captioning microservice and exercise its REST API endpoints for file upload, caption generation, streaming, chat completions, and Kafka integration.

Validates NVIDIA TAO DAFT datasets for structure, schema, and cross-reference errors using the `tao-daft validate` CLI tool.

Train, evaluate, export, and run inference for Masked Auto-Encoder (MAE) models for self-supervised pretraining and fine-tuning of visual representations.

Trains, evaluates, exports, and runs inference on FoundationStereo models for stereo depth estimation and 3D reconstruction from stereo image pairs.