agent-service-toolkit

Why we included this project

If your team is starting a LangGraph project, this toolkit spares you the usual scaffolding. It bundles a working agent definition, a FastAPI service with streaming and non-streaming endpoints, a Python client, and a Streamlit chat UI that supports voice input, so you get a complete deployable setup rather than a bare example. It also goes beyond the basics, showing human-in-the-loop interrupts, long-term memory via Store, multiple agents served by URL path, and AG-UI protocol support for connecting frontends like CopilotKit. That combination makes it a useful reference for how modern agent services are structured, and a repo you can fork and adapt instead of building from scratch.