Where cards come from

The first cards help the user reach value. They should be simple and tied to setup or first review.

Early cards should not pretend to know what the data has not shown yet.

Setup gaps include missing account context, missing files, or unclear workspace settings.

AI readiness findings can become cards when the check points to a real fix.

Missing integrations can become cards when Search Console, Google Analytics, or GitHub would unlock useful context.

Scheduled agents can send cards after prompt runs, weekly reviews, or source checks.

They should skip the card when the evidence is weak.

Users can ask CLEA to send a card from chat. This is useful when a conversation reveals one clear next step.

Turn this finding into one suggestion card.
Use a short title, one clear reason, and a CTA that points to the right dashboard view.

Analytics gaps are strong card sources. A weak prompt, repeated competitor mention, low read-to-named rate, or missing source can all become a card.

Do not create a card from a vague idea, one noisy result, or a task nobody can act on.