wool-labs / wool

What is wool?

Wool is a distributed Python runtime that lets you turn any async function or generator into a remote task with a single `@wool.routine` decorator. It runs tasks across a scalable pool of workers using peer-to-peer discovery—no central scheduler or control plane needed—handling serialization, streaming results, and exceptions transparently over gRPC. Developers get a comfyui distributed github setup or distributed python task queue without the usual orchestration boilerplate, perfect for distributed python computing on local machines or LANs.

Why is it gaining traction?

Unlike heavy distributed python frameworks like Ray or Dask, Wool skips centralized queues and coordinators for a lightweight, decentralized alternative that feels like plain asyncio. The hook is nested routines and bidirectional streaming for async generators, enabling github distributed ai pipelines or distributed mpc github workflows with zero config. It's a fresh distributed github alternative for Python devs chasing simplicity over enterprise scale.

Who should use this?

Backend Python devs building distributed task queues or python distributed semaphore systems for microservices. AI/ML engineers scaling inference like tempo distributed github or mimir-distributed github setups. Teams prototyping distributed python library integrations without Celery's overhead.

Verdict

Promising for lightweight distributed python package needs—grab it for proofs-of-concept where Ray feels bloated. With only 14 stars and 1.0% credibility score, it's early alpha: solid README and demo, but expect rough edges in production until tests/docs mature. Worth starring for woolpower in async distro.

(187 words)