Skip to main content

Testing Rules

Before you trust a rule to run on its own, test it. MailPrism lets you check a rule against real email, dig into why a message didn't match, and run a rule on demand — all without waiting for the next email to arrive.

Test against recent emails​

When you create or edit a rule, the Test Rule card lets you run it against your most recent messages without changing anything.

  1. Open a rule for editing (or build a new one).
  2. In the Test Rule card, choose Test Against Last 10 Emails.
  3. MailPrism evaluates the rule against those messages and shows a summary.

You get a count of how many emails Matched, didn't match (No Match), and were Skipped, followed by a row per email so you can see exactly which ones the rule caught.

Testing never changes your mailbox

A test only evaluates the rule — it never runs the actions. No labels are applied, nothing is archived, forwarded, or deleted. Testing is always safe.

Rules that use AI conditions​

If your rule includes AI conditions (category, urgency, sentiment, and similar), MailPrism analyzes each email as part of the test. AI analysis takes a few seconds per message, so AI-based tests run against fewer emails to stay fast. If a test runs long, some emails are marked Skipped and you'll see a notice — test again with fewer emails or simpler conditions.

Test a single email​

You don't have to test against a batch. From a rule's Analytics page, you can test one specific email against that one rule and get a detailed, condition-by-condition breakdown.

This is the fastest way to answer the question every rule-builder eventually asks:

"Why didn't this match?"​

On a rule's Analytics page, open the Not Matched tab. It lists recent emails the rule was checked against but didn't act on. For any of them, choose Test Why.

MailPrism re-evaluates that email against the rule and expands an inline breakdown showing:

  • Whether the rule Would Match or Would Not Match overall.
  • The logic in use (AND or OR).
  • Each condition with a clear pass/fail mark, the field and operator, and the value it was compared against.

That makes it obvious where a rule falls short — for example, a Subject contains "invoice" condition failing because the real subject said "Invoice #1042" and an exact-match operator was used, or an AND rule failing because one of several conditions didn't pass.

Read the failing condition, then fix it

A failed condition usually points straight at the fix: loosen an operator (equals → contains), correct a typo in the value, or switch the group logic from AND to OR. Then test again.

Run a rule now ("Execute")​

Sometimes a rule is correct — you just want it to act on an email it already skipped. On the Not Matched tab, each email has an Execute button that runs the rule against that one message for real.

  • If the email matches, the rule's actions run and MailPrism confirms how many actions were applied.
  • If it still doesn't match, you'll get a note suggesting Test Why to see the details.
Execute is not a test — it runs the actions

Unlike testing, Execute actually applies the rule's actions to that email (labels, archive, forward, and so on). Use Test Why first if you're unsure whether the rule will match.

Which tool to use​

You want to…Use
Sanity-check a new rule against real mailTest Against Last 10 Emails (rule editor)
Understand why one email was skippedTest Why (Analytics → Not Matched)
Apply a correct rule to an email it missedExecute (Analytics → Not Matched)

→ Next: Rule Library · See also Analytics & logs