Skip to content

Issue 42: Prevent duplicate document handling after upload retries

Fictional issue body; create a real issue and use its actual identifier. Project fields: Ready, Iteration I1, High priority, authorized customer worker assignee.

Outcome and value

Service staff receive one document reference for an upload attempt even if a response is lost and the client retries. Deliberate new uploads remain possible.

Acceptance criteria

  • A retry of the same attempt returns the same document reference and starts no duplicate processing.
  • Concurrent requests for the same attempt result in one document.
  • A deliberate new attempt can create a new document even for identical file content.
  • Reusing an attempt key with different content is rejected with a clear error.
  • Attempt keys are isolated between authenticated customers.
  • Operations confirms the supported retry window; expiry behavior is documented and checked.

Relevant context

Customer workspace and decision draft. Inspect the actual upload endpoint, client and existing persistence boundary.

Constraints

Reuse the existing stack. Scope this increment to retries, not general file deduplication. Product Owner decides upload intent; Technical Owner reviews design; Operations decides the retry window.

AI work status

Plan: confirm attempt semantics and retry window; inspect the existing upload boundary; implement one bounded change; verify acceptance cases; prepare a linked PR and knowledge update. Plan review: r1 draft; local review required before Build; approval evidence: none. Next action: inspect existing handling and bundle consequential questions. Evidence: none yet. Required checks and human acceptance are outstanding.

Customer-owned text and feedback outside the bounded AI section must be preserved.

Canonical source: templates/customer-workspace/examples/issue-42.md.