Audience: users learning Brain by doing.
Goal: learning. Each tutorial walks step-by-step from a known starting point to a known endpoint. You follow the instructions; the instructions work. No prior Brain knowledge assumed.
Tutorials are deliberately distinct from
../guides/: a guide says "here's how to do X in
your real environment"; a tutorial says "do exactly this and you
will end up here". Tutorials prioritise the journey; guides
prioritise the goal.
Read in order. Each one builds on the previous.
| # | Tutorial | You'll have |
|---|---|---|
| 01 | 01-quickstart-docker.md |
Brain running in Docker, health endpoint green |
| 02 | 02-first-substrate-app.md |
Memories encoded, substrate recall returning hits |
| 03 | 03-shell-deep-dive.md |
Hands-on tour of brain — encode, recall, subscribe, txn, named agents |
| 04 | 04-first-knowledge-app.md (planned) |
A schema declared, typed statements being extracted |
| 05 | 05-hybrid-query-walkthrough.md (planned) |
Hybrid retrieval with filters, query routing |
| 06 | 06-multi-shard-deploy.md (planned) |
A multi-shard topology, sized for ~1 M memories |
Tutorials are tested against each release. If a step fails:
- Check the troubleshooting section in the relevant tutorial.
- Open an issue with the tutorial filename, the step that failed,
and
docker logs brainoutput. - As a fallback, the equivalent how-to under
../guides/usually has more detail.
../concepts/— read these in parallel with the tutorials if you want background on what's happening.../guides/— the production-shaped versions of what the tutorials demonstrate.