// the find
coleam00/ottomator-agents
All the open source AI Agents hosted on the oTTomator Live Agent Studio platform!
A monorepo of AI agent examples tied to the oTTomator Live Agent Studio platform, a hosted community showcase. Each subdirectory is a self-contained agent — Python scripts, n8n JSON workflows, or Voiceflow exports — built around various patterns: RAG, knowledge graphs, web crawling, Reddit search. It's educational material first, a usable product second.
The breadth is genuine: you get working implementations of reranking, contextual retrieval, knowledge graphs, agentic RAG, and late chunking all in one place, each with its own requirements.txt and .env.example so you can actually run them. The all-rag-strategies subdirectory in particular is a decent survey of the RAG landscape with paired docs and code. Several agents include Dockerfiles and docker-compose setups, which is more than most tutorial repos bother with. The inclusion of n8n JSON workflows alongside Python means people who don't want to write code have a usable starting point.
The repo is a dump, not a library — no shared utilities, no consistent patterns, and heavily duplicated code (the same big_tech_docs corpus appears in at least two separate agent directories). The sample data is all the same set of fictional AI company docs, which makes it hard to judge whether any of these agents actually work on real content. Last push was November 2025 and the README still says 'beta' with vague promises about a growing agent library; the platform dependency is a real risk since most agents are wired to oTTomator's hosted endpoint rather than something you can run standalone. There are no integration tests for the vast majority of agents.