• The emails page now pages through your entire send history with cursor pagination - the previous build capped the list at the 100 most recent rows. Search runs server-side across recipient, subject, and sender, and uses the same query as browsing, so a message visible in the list can no longer vanish the moment you search for it
• Dashboard numbers now come from Wraps’ own record of what it sent rather than from account-wide SES metrics. Two things change as a result, and both are the count becoming correct rather than data loss. Totals no longer include mail your AWS account sent outside Wraps. And open and click counts are now unique per message - a recipient who opens the same email three times counts once, where SES counted three times - so those two numbers in particular will read lower than they did before. Account-level SES reputation is still shown, now labelled as account-wide
• Every state says what it is: a failed load says so and offers retry instead of rendering "No emails found", sandboxed AWS accounts are told they are in the SES sandbox instead of being asked to send their first email, and orgs whose event pipeline has never delivered an event see a banner explaining why everything reads as Sent
• The message detail page distinguishes an unreadable event timeline from an empty one, names the AWS account involved, and points at wraps email doctor - it previously bounced you back to the list on any failure
• Filters, search, sort, and time range live in the URL, so a filtered view survives reload and can be handed to a colleague mid-incident. Rows are real links: keyboard, middle-click, and screen readers all reach message detail now
• CSV export pages through up to 10,000 matching rows and states the cap before writing the file - it previously exported whatever was on screen and reported it as the total
• Fix: the analytics refresh button now expires the server-side cache instead of refetching the same cached payload for up to five minutes

Platform v0.21.0