Claude Skills by VectorSpaceLab

Use Dipy console entry points, workflow parser behavior, and safe CLI probes.

"Choose and run Dipy denoising and preprocessing steps, including

"Load, validate, convert, and save Dipy diffusion images,

"Choose and fit Dipy diffusion reconstruction models, compute

"Reslice Dipy images, configure affine and SyN registration, apply

"Plan and debug Dipy tractography, streamline clustering, bundle

"Use this repo skill for DiscoArt image generation,

"Generate Disco Diffusion artwork with DiscoArt Python APIs,

"Operate DiscoArt command-line, YAML execution, Jina serving,

"Use this sub-skill for DiscoArt config files, prompt schema v1,

"Routes Distil-Whisper inference, PyTorch distillation training,

"Routes Distil-Whisper JAX/Flax reproduction, long-form evaluation,

"Routes Distil-Whisper checkpoint usage, transcription, and

"Routes Distil-Whisper pseudo-labelling, student initialization,

"Use DIG (Dive into Graphs) to load graph-learning datasets, run

"Use DIG's fair graph learning stack for NBA/POKEC datasets,

"Use DIG's GOOD OOD dataset loaders for domain/shift splits,

"Use DIG's graph augmentation stack for GraphAug, S-Mixup,

"Use DIG's xgraph stack for subgraph, edge, node, walk, and

"Use DIG's large-scale graph stack for GraphFMOB/GraphFMIB-style

"Use DIG's 2D molecular graph generation stack for QM9/ZINC/MOSES

"Use DIG's graph self-supervised learning stack for GraphCL, GRACE,

"Use DIG for 3D graph learning and 3D molecular geometry

"Guide Django REST Framework and GeoDjango projects through

"Install and verify django-rest-framework-gis with GeoDjango,

"Shape GeoJSON pagination responses and OpenAPI schemas with

"Serialize GeoDjango geometries and model instances as

"Apply django-rest-framework-gis spatial filter backends, geometry

"Route DLTK 0.2.1 medical-imaging workflows across legacy

"Implement and validate DLTK TensorFlow 1.x medical-image readers,

"Use DLTK's TensorFlow 1.x prediction, full-volume sliding-window

"Select and construct DLTK 0.2.1 TensorFlow 1.x 3D networks and

"Compose DLTK TensorFlow 1.x Readers, model functions,

"Use DeepMind dm_control for MuJoCo simulation, Control Suite

"Create and validate custom dm_control Composer entities, tasks,

"Use high-level dm_control locomotion, soccer, mocap, walker, prop,

"Build, parse, compose, export, compile, step, inspect, and

"Choose and troubleshoot dm_control rendering backends, viewers,

Load, inspect, step, wrap, and validate dm_control suite environments.

"Use DeepMind dm-haiku as a JAX neural-network library for

"Use Haiku transforms to choose pure init/apply wrappers, state/RNG

"Use Haiku's optional experimental Flax interop APIs to embed Haiku

"Use this dm-haiku sub-skill for Haiku-aware JAX transform

"Route and implement Haiku model-building tasks with layers,

"Author and debug Haiku Module classes, direct parameter/state

"Use DocArray for multimodal Pydantic-style documents, typed

"Model multimodal data with DocArray BaseDoc schemas, predefined

"Serialize, persist, stream, and route DocArray BaseDoc, DocList,

"Build and troubleshoot DocArray Document Index workflows for local

"Operate doccano for text annotation, dataset import/export,