
Runs a guarded agent loop that tunes an existing NVFLARE federated job for a metric
- Open source
- 4K GitHub stars
- AI models and LLM tooling

Runs a guarded agent loop that tunes an existing NVFLARE federated job for a metric

Builds a final report and JSON summary for a stopped NVFLARE Auto-FL campaign

Converts Hugging Face Trainer or TRL SFTTrainer code into an NVFLARE federated job

Converts existing PyTorch Lightning training code into a federated NVFLARE job

Converts plain PyTorch training code into a federated NVFLARE job

Computes federated statistics over tabular or image data across NVFLARE sites

Shared NVFLARE reference files and templates that other NVFLARE skills load on demand

Drives NVIDIA Broadcast over its MCP gateway to toggle AI camera and mic effects

Collects and packages NVSHMEM put/get perftest results with system and topology evidence

Recommends NVSHMEM NIC-to-PE mappings and environment exports from GPU and RDMA topology

Plans and validates NVSHMEM and NVSHMEM4Py installs across packages, containers and source

Chooses an NVSHMEM remote transport for a target system using read-only evidence

Orchestrates CAD and source assets into SimReady USD with material, physics and validation

Diagnoses and optimizes OpenUSD stages for load time, memory and frame rate

Runs NVIDIA PAIDF Cosmos Predict to generate videos from a media sample list

Orchestrates Cosmos AnomalyGen defect image generation for PCBA, metal and glass

Sets up and validates NVIDIA physical AI infrastructure for synthetic data workflows

Guides choosing a physics engine and Newton/ovnewton integration for Physical AI

Orchestrates video data augmentation and auto-labeling runs on the OSMO platform

Checks whether a host and Jetson board are ready for NVIDIA PVA development work

Generates, changes, or repairs vectorized NVIDIA PVA operators from a validated spec

Writes a validated NVIDIA PVA operator spec, test matrix and OpGen handoff

Audits GPU quota fairness in a Run:ai cluster, finding starved and over-borrowing projects

Diagnoses why a Run:ai workload failed, crashed or stopped, with cited evidence

Checks a Run:ai cluster's node health and scheduling, with a per-pool verdict

Finds Run:ai GPUs held by projects or departments that sit idle or underused

Summarizes Run:ai workload health across a scoped set or for one workload, read-only

Drafts a governance skill card for an existing agent skill directory

Guides an agent through root cause analysis of NVIDIA TAO ChangeNet classification model

Scores object detection predictions against ground truth to produce per-class mAP metrics