Guard — Protecting Your Inbox
Guard screens every incoming customer message for hostile tone, threats, and spam before it lands in your Inbox, so you're never blindsided by an abusive customer.
Settings
- Tone Scoring - Scores each message's emotional tone (neutral, frustrated, hostile, abusive) and shows a badge in the Inbox. Includes a sensitivity level (Low/Medium/High) that controls how easily a message counts as "flagged."
- Threat Detection - Looks for five specific threat types: negative-review threats, chargeback threats, legal threats, social-media threats, and blackmail-style demands.
- Spam Filter - Catches duplicate messages, link-stuffed messages, and other spam patterns, automatically routing them to a Spam tab.
- De-escalation Drafts - When a message is hostile or threatening, generates three ready-to-send reply options (empathetic, firm, brief) for you to pick from.
- Email Alerts - Checkboxes for getting an email when a hostile message or a threat is detected, plus the address to send it to.
- Protected Phrases - A list of phrases meant to always trigger an alert regardless of tone.
- Safe List - Email addresses that skip all screening entirely — useful for known, trusted customers.
Who can use this: Guard is available identically on every plan, including Free — nothing here is a paid upgrade.
Good to know: A real threat always takes priority over spam — if a message looks like both, it's treated as a threat and never hidden in the Spam tab.
Good to know: Right now, Protected Phrases and the Email Alerts checkboxes are saved when you fill them in, but they don't currently trigger anything live — the only real-time alerts today are an internal automation trigger and a browser push notification feature that hasn't launched yet.