We see this one a lot โ an email integration (Gmail or Outlook) starts failing with a credentials error, someone reconnects the account, and the automation still fails the exact same way.
Why this happens: A failed automation run is locked to the credentials that were active at the time it ran. Retrying that same failed run doesn't pick up your newly reconnected credentials โ it's still trying to use the old, broken ones. This is easy to miss, since re-authenticating feels like it should fix everything going forward, including anything currently stuck.
The fix: After reconnecting your account, don't retry the old failed run. Instead, trigger a brand-new run โ make an actual change to a record so the automation fires fresh. That new run will use your current, valid credentials.
If it still fails after a fresh run: At that point it's worth reporting as a genuine issue, since a fresh run failing with valid credentials is a different problem than a stale retry.
Quick rule of thumb: Reconnecting credentials fixes future runs, not the one that already failed. If you're troubleshooting a credential error, always test with a new trigger, not a retry.
Run into an integration quirk like this? Tell us below.