local-agentslocal-agentslocal-agentslocal-agentslocal-agentslocal-agentslocal-agents

local-agents turns a machine at home into a small network of specialised LLM agents. An intelligent router sits in front of LM Studio and sends each request to the right one (a coder, a creative writer, and a home/utility assistant) with integrated voice and a dashboard where you assign which local model backs each role, tune the classifier, and wire up memory (embeddings) and a household knowledge base.

Everything runs on hardware you own: LM Studio serves the models on your GPU, and the router, web search, memory and dashboard run in Docker, reachable from any device on your local network. It’s the operator’s dashboard for agents that never leave the house.

Early and in progress: the router, dashboard and design-system adoption are in place; broader control is on the way. The source repository is currently private.