Claude Skills by VectorSpaceLab

"Prepare and validate KITTI or NuScenes data layouts, generated

"Use for CPU-safe SECOND box geometry, coordinate conversion,

"Route legacy SECOND and PointPillars model construction, training,

"Guides the historical SECOND KITTI web viewer, its Flask API,

"Guides SecretFlow workflows for runtime setup, federated data

"Guides SecretFlow preprocessing, statistics, and classical ML

"Guides SecretFlow component CLI, component evaluation payloads,

"Guides SecretFlow PSI, secure aggregation/comparison, Kuscia, TEEU

"Guides SecretFlow local runtime setup, device objects, and

"Use Meta Segment Anything (SAM) for prompt-based segmentation,

"Generate masks for all objects in images or folders with Segment

"Export Segment Anything's prompt encoder and mask decoder to ONNX,

"Use SAM's SamPredictor for prompt-based masks from points, labels,

"Guides segment-geospatial/SamGeo workflows for geospatial SAM

"Guides segment-geospatial's FastAPI service, samgeo-api CLI,

"Routes SamGeo SAM1 and SAM2 geospatial segmentation workflows

"Guides SamGeo geospatial utility workflows for CRS checks, tile

"Guides SamGeo3 and SamGeo3Video workflows for SAM3/SAM3.1 text,

"Routes SamGeo optional model integrations including FastSAM,

"Route segmentation_models_pytorch tasks for semantic segmentation

"Choose segmentation_models_pytorch encoders/backbones, configure

"Create and debug segmentation_models_pytorch model architectures,

"Save, reload, share, and deployment-check

"Maintain segmentation_models_pytorch source, docs, and focused

"Use segmentation_models_pytorch losses, metrics, tensor modes, and

"Use Segmentation Models for Keras/TensorFlow image segmentation

"Choose, configure, and validate Segmentation Models losses and

"Construct Keras segmentation models with Segmentation Models

"Assemble safe Segmentation Models training, evaluation,

"Guides Semantra semantic-search CLI, local document indexing,

"Guides Semantra text and PDF preprocessing, window configuration,

"Guides Semantra local web search, query arithmetic, preference

"Guides Semantra embedding model selection, OpenAI and Hugging Face

"Use Sentence Transformers for dense embeddings, semantic search,

"Use when selecting Sentence Transformers inference backends or

"Dense SentenceTransformer embedding and similarity workflows for

"Route sentence-transformers evaluation and training tasks across

"Use for sentence-transformers CrossEncoder pair scoring,

"Dense semantic search utilities, retrieve-and-rerank

"Use SparseEncoder for SPLADE-style sparse embeddings, sparse

"Routes training, inference, and Flask serving workflows for the

"Routes checkpoint inference, candidate ranking, and Flask API

"Routes checkpoint training, continuation, evaluation, and

"Use Operação Serenata de Amor's Rosie suspicious-expense pipeline

"Operate Jarbas and Rosie service setup, data loading, validation,

"Use Jarbas's Django REST API for reimbursements, companies,

"Run, inspect, adapt, and troubleshoot Rosie's

"Use SGLang for high-throughput LLM and VLM serving,

"Plan and analyze SGLang benchmarks, throughput/latency runs,

"Write and debug SGLang frontend language programs, runtime