Skip to content

Troubleshooting playbook

Collect safe evidence, stop amplification, reconcile authoritative state, and contact support.

Updated View as Markdown

Collect safe evidence, stop amplification, reconcile authoritative state, and contact support.

How it works

Collect safe evidence, stop amplification, reconcile authoritative state, and contact support. TidyMessages keeps channel-specific behavior explicit while providing stable identifiers, safe retries, and auditable lifecycle events.

Set it up

  1. Read the stable code, HTTP status, request ID, and affected resource.
  2. Classify the response as validation, permission, state, capacity, provider, media, or transient.
  3. Correct permanent failures or retry transient failures with bounded backoff and the original Idempotency-Key.

What to expect

  • The operation is traceable by stable request and resource identifiers.
  • Acceptance, provider delivery, and customer interaction remain separate states.
  • Retries and reordered events do not duplicate the customer action.

[!TIP] Apply retry budgets per resource; never replay an entire failed queue at once.

Navigation

Search for a question, a feature, or a task.

↑↓ to navigate↵ to openEsc to close