Describe what's missing in our docs
Problem
Much of our docs (task runners, secrets, storage, auth, workers, installation, config) is written for self-hosted / Enterprise, where the reader controls the infra. On Kestra Cloud the setup often differs, or the feature isn't available, and nothing on the page signals it, so readers get stuck thinking they did something wrong. A 528-page audit found ~105 pages where Cloud differs or isn't available, and only 5 note it today. (A customer recently lost ~2 days to this; see the companion AWS Batch auth issue.)
Ask
A reusable callout at the top of affected pages (extend :::alert with a cloud variant, or a shared partial, so copy + CTA are edited in one place):
- covers both "works differently on Cloud" and "not available on Cloud (EE / self-host only)";
- generic by default, with a specific one-liner + deep link where the difference is known (e.g. AWS Batch auth);
- links to a canonical "Kestra Cloud: what differs" page and offers help via (eg.
cloud-support@kestra.io -> TBD)
Example:
"Using Kestra Cloud? Some setup here differs from self-hosted / Enterprise, or isn't applicable. If a step doesn't match, it's likely a Cloud difference, not you → see [Kestra Cloud: what differs] or email cloud-support@kestra.io."
Feedback loop: each customer-reported discrepancy becomes a cloud-discrepancy doc issue, fixed case by case. This can be the Epic parent. Target pages are in the comment below.
Dependency
cloud-support@kestra.io inbox and/or dedicated Slack channel needs to be created and routed to the Cloud team before the callout ships.
Describe what's missing in our docs
Problem
Much of our docs (task runners, secrets, storage, auth, workers, installation, config) is written for self-hosted / Enterprise, where the reader controls the infra. On Kestra Cloud the setup often differs, or the feature isn't available, and nothing on the page signals it, so readers get stuck thinking they did something wrong. A 528-page audit found ~105 pages where Cloud differs or isn't available, and only 5 note it today. (A customer recently lost ~2 days to this; see the companion AWS Batch auth issue.)
Ask
A reusable callout at the top of affected pages (extend
:::alertwith acloudvariant, or a shared partial, so copy + CTA are edited in one place):cloud-support@kestra.io-> TBD)Example:
"Using Kestra Cloud? Some setup here differs from self-hosted / Enterprise, or isn't applicable. If a step doesn't match, it's likely a Cloud difference, not you → see [Kestra Cloud: what differs] or email cloud-support@kestra.io."
Feedback loop: each customer-reported discrepancy becomes a
cloud-discrepancydoc issue, fixed case by case. This can be the Epic parent. Target pages are in the comment below.Dependency
cloud-support@kestra.ioinbox and/or dedicated Slack channel needs to be created and routed to the Cloud team before the callout ships.