If you're working with Dynatrace notebooks, this builds directly on the dt-app-dashboards skill with the structural differences mapped out. Notebooks use a section array instead of tiles and grids, version "7" instead of 21, and drop variables entirely. The key workflow detail: always validate queries with dtctl first, set autoSelectVisualization to true unless the user wants something specific, and deploy through the included deploy_notebook.sh script instead of raw dtctl apply. It catches broken queries before they hit your environment. The documentation includes a clear field mapping table showing where query strings and visualization settings live in notebooks versus dashboards, which saves you from hunting through nested JSON.
npx skills add https://github.com/dynatrace/dynatrace-for-ai --skill dt-app-notebooks