Skip to main content

Tutorials

These tutorials walk you through OctoMesh from the first login to your own application. They build on each other, but each one can be read on its own.

#TutorialYou will learn
1Getting started with OctoMeshThe core concepts (tenant, Construction Kit, runtime entity, blueprint, pipeline, stream data), a tour of the Refinery Studio and how to log in with octo-cli
2Accessing dataHow to get an access token and query runtime entities, associations and time series through GraphQL, with curl and Python
3Understanding energy dataA use case end to end: the data model of an energy community, metering registers (OBIS codes), data quality, the daily allocation and the key figures derived from it
4Building solutions with AI via the MCP serverHow to connect an AI assistant to OctoMesh, explore data in natural language and let the assistant generate an evaluation or a small app
5Bringing your own data via pipeline endpointsHow HTTP endpoints on the Mesh Adapter work and how to call them from your own devices or services

What you need​

  • An OctoMesh tenant and a user account in it. The examples use the placeholders <tenant> for the tenant id and <your-domain> for the domain of your OctoMesh installation, so the asset repository is reachable at https://assets.<your-domain> and the identity service at https://connect.<your-domain>. Your administrator tells you the actual values.
  • A web browser for the Refinery Studio.
  • For the command-line parts: octo-cli, curl, and Python 3.10 or newer with the requests package.
  • For tutorial 4: an MCP-capable AI client such as Claude Code or Claude Desktop.
Read-only first

Tutorials 1 to 4 only read data. You can follow them in a productive tenant without changing anything. Tutorial 5 writes data, so use a test tenant for it.

Where to go next​

The tutorials link to the reference documentation wherever a topic goes deeper: