Persistent state fabric.

Refs over a KV backend (RocksDB, LMDB). Transactions, snapshots, and change notifications, built in.

Web UI fabric.

Refs are widgets: text, buttons, tables. The fabric renders them in the browser and live-updates as your state changes.

Cluster compute fabric.

Teleport a Nu tree to any worker in your cluster; it runs there and returns the result. Backed by Ray under the hood.

LLM chat fabric.

One OpenAI-compatible wire, N providers. Ollama, OpenAI, OpenRouter, Groq, Cerebras, xAI, vLLM — same ChatRef, same call.

In-memory state fabric.

In-memory state on plain dicts. Perfect for cache, hot state, and in-process coordination.

Proxy fabric.

Puts other fabrics on the network. Bind a fabric in one process, use it from another; same Refs, over TCP or Unix socket.

HTTP fabric.

Expose Nu Refs as HTTP endpoints, or build fabrics on top of any HTTP service. Nu meets the web.

In-process service fabric.

Wrap any Python object as a Nu Service. Its methods become Refs — queries, actions, streams, commands — you compose into the tree.

Claude Code fabric.

Claude Code as a Ref. Prompt from your Nu tree, scope sessions with a bracket, get text and metadata back.

Multiprocessing fabric.

Run Nu trees in parallel across local processes. Backed by Python multiprocessing — no cluster required.

Worker pool fabric.

The pool itself is the fabric. One Provide owns N worker processes, so you launch one, kill one, or run a tree on the worker you name.

Websocket server fabric.

Run a websocket server from your Nu program. Every client that connects gets its own arm of the tree, and you decide what travels over the wire.