WIP (2026Q2)
This is a new repository built around a first-class observability initiative for Materialize.
As of Materialize 27.0 (2026Q2), this repository is still under active development and not supported.

Dashboards as Code#

Instead of more common ClickOps strategies (manually configuring dashboards in the UI), we manage dashboards as reproducible source code. Sources live under packages/grafana-dashboards/.

The audience for this section is repo contributors — SRE, Field Engineering, CloudOps, Database Engineers — and AI agents reading the corresponding dashboards-as-code skill. The audience for the dashboards themselves (panel descriptions, naming, visual choices) is the Materialize end user; see Style Guidelines for that voice.

In this section#

  • SDKs and Schemas — Grafana versions we target, Dashboard v1 vs v2 schema state, and the grafana-foundation-sdk / py-mzmon-lib toolchain.
  • Style Guidelines — palettes, layouts, panel visualization conventions, description voice, PromQL conventions, Materialize metric label families, known gotchas, and PromQL recipes.
  • Generating and Pushing Dashboards — code structure, UID determinism, the __main__ entry point, the production push path (gcx dashboards update), and the ad-hoc Grafana v2 API path for iteration.
  • Testing — testing conventions (currently sparse).

The dashboards-as-code Claude skill at .claude/skills/dashboards-as-code/SKILL.md is the live state snapshot — current dashboard inventory, in-flight stubs, and cleanup candidates — and links back into the pages above for stable reference.

Backward Skills SDKs and Schemas Forward