n8n Workflow: Restore Supabase Row Updates via Rewind HTTP API

Restore Supabase Row Updates Using Rewind’s Rewind HTTP API (with Conflict Protection)

This n8n workflow claims an approved Rewind recovery job and conditionally restores a Supabase row update—only when the stored revision still matches. If a later edit exists, it preserves the newer data and reports the result back to Rewind.

What this workflow does

• Runs during setup (manual) to validate the end-to-end recovery process before scheduling it.
• Validates configuration, including the two service origins and the connection name used for the Rewind capture.
• Claims one human-approved recovery job from Rewind.
• Validates the supported update, table alias, and row ID included in the job.
• Reads the current Supabase resource and compares:
  • the current revision vs. the recorded revision from Rewind
  • the recorded fields captured at the time of the Rewind capture
• Conditionally restores the saved fields only when revisions match; otherwise it preserves newer edits.
• Reports the outcome to Rewind, then you independently verify the restored row directly in Supabase.

Use cases

• Recovering from accidental or harmful updates to a Supabase row after Rewind has captured the event and a human approved the recovery.
• Preventing data clobbering by ensuring a restore only happens when no newer edits occurred since capture.
• SaaS operations and automation engineering workflows where recovery steps require approval and auditability.

Technical details

• n8n nodes: if, set, code, sticky note, http request, and manual trigger.
• Integrations:
  • Rewind HTTP API via “Claim approved job” and “Report outcome” using Rewind worker Header Auth.
  • Supabase via Read current resource and Conditional restore using Supabase apikey Header Auth.
• Prerequisites: a Rewind account, a registered Supabase row captured with Rewind, installed connector RPCs, and separate worker/server credentials stored in n8n.