vmatiyko —
Hi everyone! 👋 I left a proposal on #3796 about handling `connector_sync` cancellation before considering a broader `pg-boss` migration. My read is that connector deletion needs DB-visible cancellation either way, because running sync handlers still need cooperative checkpoints before ingesting more data, enqueueing embeddings/continuations, or marking success. Would appreciate a quick sanity check on whether this direction matches what you had in mind before I start implementing.
