Claude Skills by VectorSpaceLab

"Prepare and validate datasets for MMYOLO:

Guide MMYOLO export, checkpoint conversion, and deployment backend workflows.

"Guide MMYOLO image/video inference, LabelMe export, large-image

"Inspect and extend MMYOLO registries, model components, datasets,

"Route safe MMYOLO training, testing, prediction output, launchers,

"Route MNE-Python neurophysiology I/O, preprocessing,

"Guides MNE-Python CLI entry points, dataset helpers,

"Read, create, inspect, annotate, concatenate, save, and export

"Guide MNE-Python event extraction, sensor-space preprocessing,

"Guide safe MNE-Python repository maintenance and editing while

"Guide MNE-Python source-space, BEM, forward, covariance, inverse,

"Guide MNE-Python time-frequency, spectral, statistics, decoding,

"Guide MNE-Python plotting, visualization backends, topomaps,

"Guide MOABB EEG and BCI benchmarking workflows: choose datasets,

"Analyze MOABB result DataFrames or Results stores: validate score

"Select, inspect, configure, and safely load MOABB datasets,

"Choose a MOABB generalization protocol and splitter, then run

"Choose a MOABB task paradigm and build a verified preprocessing,

"Operate the MobileAgent repository family: GUI-Owl/Mobile-Agent

"Prepare and troubleshoot MobileAgent GUI benchmark and evaluation

"Use current Mobile-Agent-v3.5 / GUI-Owl 1.5 phone, desktop, and

"Preserve, run, and migrate legacy Mobile-Agent v1, v2, and v3

"Use Mobile-Agent-E individual and self-evolution Android

"Use PC-Agent current and v1 desktop automation workflows with

"Prepare UI-S1 / verl GUI trajectory post-training, SOP evaluation,

"Guides modAL-python active-learning workflows, query strategies,

"Guides modAL BayesianOptimizer and PI/EI/UCB acquisition workflows

"Optional DeepActiveLearner, PyTorch/skorch MC-dropout, and

Router for ActiveLearner, Committee, and CommitteeRegressor workflows.

"Router for modAL uncertainty, disagreement, batch, density,

"Use ModelScope SDK workflows for Hub access, pipelines, datasets,

"Extend ModelScope with custom pipelines, registered components,

"Operate ModelScope dataset loading, file IO, and config parsing

"Use ModelScope Hub, cache, download/upload, and CLI workflows

"Use ModelScope inference pipelines, model/preprocessor registries,

"Serve ModelScope pipelines, plan vLLM handoffs, use exporter APIs,

"Use for ModelScope trainer construction, TrainingArgs conversion,

"Guide Modin pandas-compatible, distributed dataframe, I/O, engine

"Use Modin experimental Batch Pipeline, XGBoost, spreadsheet,

"Use Modin's pandas-compatible DataFrame and Series API safely and

"Configure Modin engines, backends, resources, diagnostics, and

"Use Modin readers, writers, conversion APIs, glob I/O, SQL

"Use MolecularNodes 5.2 in a Blender 5.2 host to import, style,

"Operate MolecularNodes density-grid and ensemble workflows in a

"Load static MolecularNodes structures, compose geometry-node

"Configure MolecularNodes Blender scenes, cameras, engines,

"Install and verify MolecularNodes in a Blender 5.2 host, then

"Operate MolecularNodes Universe-backed trajectories: map Blender

"Use MONAI, the PyTorch-based medical imaging AI toolkit, for

"Use high-level MONAI application workflows, especially Auto3DSeg