Reference pages are information-oriented: the facts, laid out for looking
up. One page per component, mirroring templates/<name>.yml 1:1. If you
want a guided lesson instead, see Tutorials; for a
task-oriented recipe, see How-To Guides; for the why
behind a design, see Explanation.
Which pipelines each component runs on —
the trigger matrix for all 31 components (merge request, branch, default
branch, tag, schedule), which triggers you can override, and why a
component that appears to have been ignored usually produced no job at
all.
Limitations — what these components deliberately
do not do: unsupported platforms and clouds, combinations that do not
work, and the things you have to supply yourself.
Runs a project's own documentation check on every merge request. Deliberately always-on: a docs gate filtered on docs paths skips the merge requests that break the docs.
Scans a built image with trivy — OS packages as a hard gate, tool binaries as advisory — and keeps one dashboard issue per image current with the findings.
Schedule-driven Renovate in autodiscover mode — one job manages every already-configured repo under a group filter, picking up new repos automatically.
Derives the estate's first-party dependency order from its go.mod files and plans (or walks) the release cascade in it — the sequencing Renovate structurally cannot do.