Stop it. Reload it. Resume it.
Run a fixed two-record workflow in this page. The first run saves one record and stops. The second run reads that checkpoint, avoids fetching the completed record again, finishes the remaining record, and skips a duplicate.
01 · RUN
Save one record
A-001 is written before the simulated stop.
02 · RELOAD
Checkpoint survives
You can reload this page before resuming.
03 · RESUME
Finish without refetch
A-001 is skipped before fetch; B-002 is completed.
Ready — start Run 1.
Event timeline
- Waiting for Run 1.
Checkpoint and proof
Durably completed IDs
[]
Forced stop observed
Completed records refetched
Final unique records
Duplicates skipped
Generated result
0 unique records
| ID | Title | Price | Source |
|---|---|---|---|
| Run the demo to generate rows. | |||
Scope: this is a fixed offline browser walkthrough, not the paid Python/Playwright engine and not performance on arbitrary websites. It demonstrates the same narrow state transition with fixed sample data. Use the free Python lab below for the real process-exit proof.