Skip to content

Built-in checks

Built-in checks catch mistakes every company makes, such as an email going to the wrong person, so you don’t have to write a policy for them. They run on the same emails as your policies, and their warnings appear the same way: an underline under the words, and a private card when you click it.

Check What it catches Runs on
Recipient check Email going to, or written about, the wrong person Gmail, Outlook on the web
Consistency check A detail that doesn’t match the thread or the calendar Gmail, Outlook on the web
Fact check A price, limit or date that differs from your own pages Gmail, Outlook on the web

Each check is made of parts that you can turn on or off one at a time. For example, the recipient check has three: greeting names, writing about a recipient, and Reply vs Reply all.

Open Checks from the sidebar. The page lists every check your workspace can run, with an example for each part.

  • Who can see it: everyone in the workspace.
  • Who can change it: admins (the MANAGE_CHECKS permission, see Roles & permissions). Everyone else sees the switches but can’t use them.
  • Per part or all at once: each part has its own switch, and the switch next to a check’s name turns all of its parts on or off together.
  • Default: every part of every check is on until an admin turns it off.
  • History: each change is recorded in Activity Logs with who made it and what changed.

A check shows on the page only while it can run for your workspace. If no check can run, the Checks page and its links are hidden, and opening the page sends you back to the dashboard.

Built-in check cards appear in the email compose window, next to any policy cards, and look the same apart from three things:

  • The card’s header names the check (for example “Recipient check”) instead of “Policy match”.
  • There is no link to a policy, because no policy is behind the warning.
  • Some cards offer a fix (for example, the right spelling of a name). Others only explain the problem.

As with policy cards, the card is visible only to the writer, nothing is blocked, and sending is always the writer’s call. Clicking 👎 hides that warning for that writer from then on.

The Analytics dashboard shows built-in checks in their own Built-in checks panel, one row per part, with the cards shown, fixed, self-resolved and ignored over the selected period. A misspelt greeting is not a policy violation, so these cards are not counted in the policy figures (the KPI row, Top triggered policies, and the area and severity charts). See Metrics reference.

The same privacy rules apply as for policies: events carry the check and the outcome, never the email text or who wrote it.

The fact check is not in this panel yet.