Quickstart (Ollama)

This tutorial runs RexOS locally using Ollama’s OpenAI-compatible endpoint.

Prerequisites

• Ollama is installed and running.
• You have at least one chat model available (example: qwen3:4b, llama3.2, etc.).

1) Start Ollama

ollama serve

2) Initialize RexOS

This creates: - ~/.rexos/config.toml (provider config + routing) - ~/.rexos/rexos.db (SQLite memory)

rexos init

3) Run your first agent session

Pick a workspace directory (tools are sandboxed to this root):

mkdir -p /tmp/rexos-work
rexos agent run --workspace /tmp/rexos-work --prompt "Create hello.txt with the word hi"

RexOS prints the final assistant output, and also logs a session_id to stderr for later reuse.

4) Re-run with the same session id (optional)

rexos agent run --workspace /tmp/rexos-work --session <SESSION_ID> --prompt "Now append a newline + bye to hello.txt"

Next steps

• Use the harness for long tasks: see “Long Task With Harness”.
• Switch providers/models: see “Providers & Routing”.