Hi community,
Durable Execution is here to make flows more reliable — even during deployments! 
Why It Matters:
Before, deploying new features or fixes could stop running flows.
The server would warn the evil engine:
“You have 30 seconds to finish, or you’ll be shut down!”
The engine tried to finish, but flows often needed more time.
When unable to complete in time, the queue restarted the run from the beginning.
This caused some steps to run twice and made execution unreliable. 
What’s New:
Deployments now work better. The server gives a new message:
“You have 30 seconds to finish and upload all progress so far.”
Progress is now saved, so after deployment, the engine continues from the last saved step.
No more repeated steps or lost data.
The worker also saves state every 15 seconds. 
What Changes:
- Duplicate executions during deployments — gone.
- Flows are now more stable and reliable.
Excited to share this update — feedback is always welcome! 
