`stripe_worker` not checking `sync_job` for error status, not syncing for 7 days after an error occurs.

[aergener]

Bug report

• I confirm this is a bug with Supabase, not with my own application.
• I confirm I have searched the Docs, GitHub Discussions, and Discord.

Describe the bug

I just created a new Supabase project, a new Stripe account, and set up Stripe Sync Engine using the integrations UI on Supabase. The first sync job in the stripe_worker ran successfully for 9 tables, but failed for 9 tables due to throttling. The sync_runs view has the status error and the following error_message:

Initialization failed: Error: Request rate limit exceeded. You can learn more about rate limits here https://stripe.com/docs/rate-limits.;

Despite the job failing, whenever the stripe_worker Edge Function runs, it prints out a message such as:

Skipping resync — a successful run completed at 2026-03-28T13:54:11.659Z (within 604800s window). Cron paused until 2026-03-28T17:58:00.764Z.

This behavior is incorrect, as the run was not successful. When I look at the Edge Function code in my Supabase account, I see the following code:

const completedRun = await stripeSync.postgresClient.getCompletedRun(
  stripeSync.accountId,
  SYNC_INTERVAL
);
const message = `Skipping resync \u2014 a successful run completed at ${completedRun?.runStartedAt.toISOString()} (within ${SYNC_INTERVAL}s window). Cron paused until ${skipUntil}.`;

...

async getCompletedRun(accountId, intervalSeconds) {
  const result = await this.query(
    `SELECT r."_account_id", r.started_at
      FROM "${this.syncSchema}"."_sync_runs" r
      WHERE r."_account_id" = $1
        AND r.closed_at IS NOT NULL
        AND r.closed_at >= now() - make_interval(secs => $2)
      LIMIT 1`,
    [accountId, intervalSeconds]
  );
  if (result.rows.length === 0) return null;
  const row = result.rows[0];
  return { accountId: row._account_id, runStartedAt: row.started_at };
}

This clearly shows that the code is not checking for the status of the sync job, it's only checking if a run completed within the last 7 days. In fact, the _sync_runs table it's checking doesn't even have a status column, that's only in the sync_runs view.

This leads to a problematic state where the stripe_worker will never error correct as long as any run has happened in the last 7 days.

To Reproduce

Steps to reproduce the behavior, please provide code snippets or a repository:

1. Create a new Supabase project, a new Stripe account, and set up Stripe Sync Engine using the integrations UI on Supabase.
2. Get throttling errors (or any other service errors) from Stripe.
3. stripe_worker will now skip syncing for 7 days.

Expected behavior

stripe_worker should check the status of the sync job, and if it's in an error state, it should run the sync job.

Screenshots

sync_runs table

stripe_worker logs

stripe_worker code

System information

• OS: macOS
• Browser: safari
• Version of supabase-js: N/A
• Version of Node.js: N/A

Additional context

getCompletedRun function
https://github.com/stripe/sync-engine/blob/main/packages/sync-engine/src/database/postgres.ts#L768

reconciliationRun comment says it should "run only if no run completed successfully in the last 24 hours", but relies on the getCompletedRun function, which only checks for a completed run, not a successfully completed run.
https://github.com/stripe/sync-engine/blob/main/packages/sync-engine/src/database/postgres.ts#L787