Skip to main content
Corrective retry changes course on purpose. Use it when automatic recovery is not enough and evidence proves a real correction is required. Ordinary document failures still need corrected evidence first. Corrective retry only continues when policy and proof allow it.

​
How it works

Lambda treats the original CargoWise capture as immutable accounting evidence. Before anything mutates, it asks:
  1. Is this document already successful, cancelled, in flight, or waiting on a parent?
  2. Does policy allow a stamp retry for the active classification?
  3. Can we resume the exact stored attempt, or must we reprocess from the original capture?
  4. For failed documents: is there proof the fiscal projection changed, or that an approved prerequisite was corrected?
If any answer fails, retry blocks.

​
Resume vs reprocess

Lambda chooses the strategy from evidence. People do not pick it. If Facturapi already created the invoice, Lambda retrieves and finalizes that invoice. It does not create again. If the create outcome is unknown, Lambda resumes the exact request. Remapping uncertainty into a new attempt is blocked.

​
Preview, then apply

  1. Preview recomputes what would happen and returns a plan fingerprint.
  2. Apply recomputes under lock and only proceeds if that exact plan still matches.
  3. A private reason is stored on the ops run. It is not copied into customer notifications.
Stale previews fail closed.

​
Hard blocks

Corrective retry never:
  • Edits the original CargoWise capture
  • Retries successful or cancelled documents as an everyday action
  • Stacks a second stamp lane on in-flight or parent-waiting work
  • Opens a new attempt when the provider may already have created the invoice
  • Retries an unchanged rejected request without proven correction
  • Uses delivery retry to issue another CFDI
Retry CargoWise delivery only re-sends XML/PDF (or a cancellation receipt). The CFDI may already be issued. Documents observed as Stamped externally have no Lambda stamp retry here. They are recorded evidence, not Lambda-owned CFDIs to remake.

​
What you do in the workspace

Support-owned corrective retry follows the same preview → apply gates when the workspace correctly blocks day-to-day action (for example after a proven mapper fix). Replacement of an already-issued Lambda-owned CFDI is a separate, explicit support operation — not the same as resume or reprocess above.