Superagent
Open-source framework for production AI assistants.
About
Superagent is an open-source framework for building, deploying, and managing LLM-powered AI assistants with APIs. Developers configure agents with data connectors, retrieval over vector stores, third-party API tool calls, memory, streaming, and cron/webhook triggers, then invoke them through REST endpoints or the Python and TypeScript SDKs. Agents can run on Superagent's cloud or be self-hosted.
The project is MIT-licensed, was backed by Y Combinator (W24), and has gathered around 5,800 GitHub stars. Typical uses span customer-support chatbots, code review bots, legal document assistants, content generation, and data aggregation pipelines. Documentation lives at docs.superagent.sh; note the npm/PyPI SDKs show older published versions, so check the repo for current development state.
What it can do
- Assistant builder with RAG, memory, and tool calls\nREST API plus Python and TypeScript SDKs\nSelf-hosted or managed cloud deployment\nWebhooks, cron triggers, and third-party API connectivity
Setup
Self-host from github.com/superagent-ai/superagent or use the hosted cloud
Prerequisites
A model provider API key; Docker for self-hosting
Use cases
Works with
Trust score — 8/10
Scored 8/10 (verified tier, metrics-v1). Evidence: Verified via vendor site/docs/GitHub and reputable coverage, wave 101; Trending score 76/100 from researcher assessment.
★ 5,811 GitHub stars
Sources
Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.