Cited answers · EU-hosted · sovereign
AskOpen Source Europe
A public assistant that searches the official OSE documentation and answers with sources you can verify. Fiscal hosting, onboarding, policies, processes — ask in your own words.
Use it in your own tools — or run it locally
The OSE knowledge RAG speaks MCP. Wire it into your editor, install the Claude Code plugin, or clone the repo and run this whole chat on your machine.
{
"mcpServers": {
"ose-knowledge": {
"type": "stdio",
"command": "uvx",
"args": ["ose-knowledge-mcp==0.1.1"]
}
}
}One package, zero install — the whole knowledge RAG runs on your machine.
/plugin marketplace add opensourceeurope/ose-knowledge-mcp
/plugin install ose-knowledge@aiMore than the MCP server — the plugin also ships an ose-researcher subagent tuned to query the knowledge RAG efficiently and return concise, cited answers.
git clone https://github.com/opensourceeurope/ose-knowledge-mcp.git
cd ose-knowledge-mcp
ollama pull mistral-small3.2 # the default local model
./scripts/run-local.shNeeds uvx, npm, python3 and a running Ollama. Starts the MCP server, the agent function and this chat UI on localhost — no API keys, nothing leaves your machine. It uses mistral-small3.2 by default; pull it first (the script won't download it for you). Already have another model? Point it at that instead — e.g. OLLAMA_MODEL=llama3.1:8b ./scripts/run-local.sh.
Sovereign by design
The entire request path runs on European infrastructure — from the moment you ask a question to the model that answers it. Nothing leaves the EU.
Answers are generated by Mistral Small (mistral-small-latest) — a model built by the European provider Mistral and run on its EU platform.
The static page, the serverless function and the MCP knowledge server all run on Scaleway (EU).
Your conversation lives only in your browser's memory. Ticking the analytics box simply turns on logging of the text you type — we drop it into a standard storage bucket hosted in the EU, where we can read and analyse it to spot gaps in the docs. Nothing else is stored.
Try:
This chat is a public demo of the OSE knowledge RAG, running a small default model. For the best answers — and as the recommended setup for day-to-day work with Open Source Europe — connect the knowledge RAG to your own agent and LLM, almost certainly more capable than the model here. See Local-first to integrate it.