Claude Skills by VectorSpaceLab

"Maintain and migrate langchain-classic legacy chains, retrievers,

"Work on langchain-core primitives: runnables, messages/content

"Work on LangChain partner integration packages under libs/partners

"Navigate LangChain's Python monorepo packages, package-local uv

"Select safe LangChain package verification commands, standard

"Use LangExtract to extract structured data from unstructured text,

"Build robust LangExtract extraction pipelines with examples,

"Create, package, register, test, and troubleshoot custom

"Choose, configure, inspect, and troubleshoot LangExtract model

"Use LangExtract visualization to save, reload, inspect, and render

"Use Langflow's visual AI workflow builder, FastAPI backend, LFX

"Work on Langflow FastAPI backend routes, graph execution,

"Create and maintain Langflow Python components, extension bundles,

"Install, run, configure, deploy, and troubleshoot Langflow in

"Use and maintain lfx, lfx run, lfx serve, lfx-mcp, Flow DevOps

"Build, validate, import/export, tweak, and run Langflow flow JSON

"Modify Langflow's React/Vite frontend, graph workspace, stores,

"Maintain the Langflow monorepo safely across tests, formatting,

"Integrate with Langflow through REST APIs, OpenAPI contracts,

"Build, run, persist, deploy, and operate LangGraph Python

"Create, validate, run, containerize, and troubleshoot LangGraph

"Build, compile, invoke, stream, debug, interrupt, and route custom

"Choose and wire LangGraph persistence: checkpoint savers, stores,

"Use LangGraph prebuilt ReAct agents, ToolNode, ValidationNode,

"Use LangGraph Python SDK clients, API resource clients, auth

"Use Langroid to build Python LLM agents, tools, tasks,

"Build and debug Langroid ChatAgent, Task, and ToolMessage

"Use Langroid structured-data agents for table/CSV/Pandas, SQL,

"Connect Langroid agents to MCP tools, web/search/file tools,

"Configure direct LLM, embedding, and model-provider access for

"Document ingestion, parsing, chunking, embeddings, vector stores,

"Route LatentSync video generation, raw-video preprocessing,

"Convert raw videos into the training-ready, face-aligned,

"Route LatentSync scoring, comparison, and diagnostic plotting for

"Run LatentSync single-pair, small-batch, Gradio, and

"Configure and launch LatentSync U-Net and SyncNet training with

"Guides LaTeX-OCR pix2tex image-to-LaTeX OCR, CLI, Python API, GUI,

"Guides pix2tex formula extraction, tokenizer and dataset pickle

"Guides LaTeX-OCR GUI, FastAPI, Streamlit, and Docker service

"Guides pix2tex CLI and Python OCR inference from equation images

"Guides pix2tex model training, resizer training, evaluation, YAML

"Routes LayoutParser workflows for document layout objects,

"Routes LayoutParser JSON, CSV, dataframe, PDF, and

"Routes LayoutParser model-zoo inference, backend selection, and

"Routes LayoutParser coordinate primitives, layout containers,

"Routes LayoutParser OCR wrappers, response parsing, and text

Routes LayoutParser box and text rendering workflows on images.

"Guides LazyLLM low-code LLM application workflows including

"Guides LazyLLM function-call tools, ToolManager, agents, sandbox

"Guides LazyLLM installation, optional extras, CLI commands,