Claude Skills by VectorSpaceLab

Load CLIP models, prepare images/text, and run safe image-text inference.

"Build CLIP zero-shot class prompts, tokenizer-safe templates, and

"Routes tasks for discovering NLP datasets in the CLUEDatasetSearch

"Guides text classification, topic classification, sentiment,

"Guides summarization, machine-translation, broad corpus,

"Guides semantic matching, paraphrase, similarity, textual

"Guides question-answering and reading-comprehension dataset

"Guides named-entity-recognition and sequence-labeling dataset

"Use code-review-graph for local-first CLI/MCP code knowledge

"Explore code relationships, search, flows, communities,

"Install, configure, update, serve, and troubleshoot

"Handle code-review-graph custom languages, embeddings, wiki,

"Review changed code, blast radius, test gaps, and PR comments with

"Routes CogDL graph-learning workflows for experiments, graph data,

"Routes CogDL experiment() and training-CLI workflows for

"Use CogDL graph data objects, custom datasets, built-in dataset

"Routes CogDL model registry, layer, custom GNN, and sparse

"Routes CogDL pipeline apps, embedding generation, recommendation,

"Routes CogDL Trainer, wrapper matching, configuration, and

"Routes Cognee AI memory platform tasks across SDK memory APIs,

"Guides Cognee custom graph schemas, ontologies, custom pipelines,

"Guides Cognee session memory, typed memory entries, feedback,

"Operates Cognee through the CLI, FastAPI app, Docker Compose

"Configure Cognee runtime, optional extras, provider settings,

"Guides Cognee's primary memory workflows: remember, recall,

"Helps users choose Cognee search and recall modes, tune retrieval

"Use ColabFold for protein structure prediction workflows: validate

"Plan and run ColabFold batch structure prediction workflows,

"Prepare and validate ColabFold FASTA, CSV, A3M, AlphaFold3

"Generate ColabFold MSAs using the public server, local MMseqs2

"Post-process ColabFold results with Amber/OpenMM relaxation,

"Use ColBERT/colbert-ai for late-interaction retrieval: prepare

"Optional ColBERT Baleen multi-hop retrieval extension: HopSearcher

"Use when preparing or validating ColBERT

"Mutate existing ColBERT indexes with IndexUpdater, persist index

"Build ColBERT indexes, load existing indexes, run single-query and

"Use when inspecting ColBERT checkpoints, ColBERTConfig

"Train or fine-tune ColBERT models, prepare and validate triples,

"Use richzhang/colorization PyTorch colorizers for automatic image

"Run ECCV16 and SIGGRAPH17 automatic image colorization to saved

"Programmatic Python API for the colorizers package: constructors,

"Use ColossalAI for distributed PyTorch training, large-model

"Navigate ColossalAI first-party application packages, environment

"Use ColossalAI Booster, plugins, dataloaders, checkpointing, LoRA,

"Use Colossal-Inference APIs and commands for LLM generation,

"Install ColossalAI, verify PyTorch/CUDA and optional extensions,

"Plan ColossalAI tensor, pipeline, sequence, hybrid parallelism,

"Use ComfyUI as a modular AI content-creation engine: launch and

"Implement, inspect, test, and package ComfyUI custom nodes,

"Configure ComfyUI model folders, extra model paths, supported