Claude Skills by VectorSpaceLab

"Build, inspect, mutate, serialize, and debug Cirq circuits and

"Use Cirq provider packages, provider-safe serialization, and

"Run local Cirq simulator, sampler, sweep, histogram, state,

"Use Cirq transformers, target gatesets, and routing workflows to

"Use the city2graph 1.0.0 Python package to construct, validate,

"Use city2graph Overture Maps acquisition, boundary resolution,

"Convert city2graph GeoDataFrame graph tables to and from NetworkX,

"Build spatial OD mobility graphs and local GTFS/GBFS

"Build and post-process spatial graphs with city2graph proximity,

"Build projected-CRS urban morphology graphs from building

"Use CLAM for weakly supervised computational pathology workflows

"Convert CLAM patch coordinate files into feature .h5 and .pt

"Configure, validate, run, and troubleshoot CLAM attention heatmap

"Prepare CLAM dataset CSVs and splits, train CLAM or MIL models,

"Prepare CLAM whole-slide-image masks, coordinate patch bags,

"Use ClawBio for local-first bioinformatics agent workflows:

"Operate the ClawBio installation, CLI, public Python runner,

"Choose a ClawBio skill or short skill chain from catalog metadata,

"Operate ClawBio's optional MCP stdio server, messaging and web

"Create, register, test, package, and maintain ClawBio skills and

"Validate ClawBio runs, benchmarks, reproducibility records, action

"Routes cleanlab data-centric AI workflows for finding label

Route standard multiclass and binary noisy-label workflows for cleanlab.

Route end-to-end dataset auditing with cleanlab.Datalab.

"Routes unstable cleanlab experimental helpers for low-memory

"Route cleanlab workflows for consensus labeling, annotator

Route cleanlab outlier and OOD scoring from features or pred_probs.

"Route cleanlab token classification, object detection, and

"Route multi-label classification and regression label-issue

"Use CleanRL's single-file reinforcement learning scripts,

"Evaluate CleanRL saved models, inspect local artifacts, and safely

"Operate CleanRL experiment matrices, tracking, tuning, resume,

"Maintain CleanRL-style code, docs, tests, metadata, and

Choose, run, inspect, and safely adapt CleanRL single-file training scripts.

"Route ClearerVoice-Studio tasks for ClearVoice inference,

"Operate pretrained ClearVoice inference for speech enhancement,

"Choose, validate, and run SpeechScore objective metrics for

"Guide ClearerVoice-Studio training, fine-tuning, launchers, data

"Route ClearML Python SDK tasks across experiment tracking,

"Build and debug ClearML automation workflows with

"Use ClearML Dataset, StorageManager, clearml-data CLI, storage

"Instrument Python ML and data scripts with ClearML Task experiment

"Configure ClearML and build safe remote execution commands or SDK

"Use ClearML HTTP routers, local FastAPI proxies, endpoint

"Guides CLIP-as-service client, server, and CLIP search workflows

"Guides clip_client.Client connectivity, encoding, ranking,

"Guides CLIP-as-service search flows with CLIPEncoder plus AnnLite

"Guides clip_server installation, Jina Flow YAML,

"Use OpenAI CLIP for zero-shot image-text inference, prompt

Extract CLIP features and evaluate downstream workflows safely.