Morning Slack digest of your Fivetran pipeline health

Every weekday morning at 8am, give me a health check of all my Fivetran data pipelines and post it to my data team's Slack channel so they start the day knowing exactly what needs attention. Use a cron trigger set to Monday through Friday at 8am.

First, use Fivetran's List Connections operation to retrieve every connection in the account. Fivetran wraps the results under data.items, and each connection carries a status object. Read its sync_state and setup_state, the last successful sync timestamp (succeeded_at, and failed_at when present), and whether the connection is paused. One call enumerates the health of the whole account, so page through with the cursor if there is more than one page of results.

For any connection that looks broken or ambiguous, you can call Get Connection or Get Connection State to pull extra detail before writing it up.

Then write a short, skimmable digest grouped by status, ordering broken pipelines first. Lead with any connection that is broken or failing, meaning a sync_state of error or broken, a failed last sync, or a setup that never finished. Next, call out anything that looks stale or paused, such as paused connections and connections whose last successful sync is well past their normal schedule (for example, no successful sync in over a day). Finally, confirm the rest are healthy with a brief count. For each flagged pipeline, name it by its source and destination or schema name, say what is wrong, and note when it last synced. Keep it scannable with clear status groupings and short lines.

Send the finished digest to the data team's Slack channel using Slack Bot's Send a Message operation. Format it with Slack mrkdwn so the status groups and pipeline names are easy to scan, and keep the wording plain by saying 'data pipelines' rather than 'connectors'.