Cold pitches, bounced
AI screens every stranger who writes to you. Pitches go to a pile you review when you feel like it, and real people always get through.
Features
A team inbox, how a team works it together, the API your product sends with, and the deliverability work that makes both of them arrive. Plus the agent layer: mailboxes for agents, an injection shield, an autonomy dial and one-click recipes. Everything on these pages is live. What is not is listed on each page.
Your team's mail, on your domain, with the parts nobody enjoys handled before you open it.
Read moreOne JSON POST, or the SMTP relay if your code already speaks SMTP. Replies come back to a human.
Read moreThe unglamorous half of email. Set up once, checked live, and enforced afterwards.
Read moresupport@ and sales@ belong to the team. Team features make it clear who has each conversation, without a ticket system on top.
Read more[email protected] is a real mailbox. The agent works from it through the API or MCP, and your team can open the same conversation in the inbox.
Read moreAnyone on the internet can write to your agent. The injection shield checks what they wrote before your agent reads it.
Read moreAgents that cannot do anything are not useful, and agents that can do anything are not safe. The dial sits in between, and you set it per agent.
Read moreRecipes are agents someone already thought through. Install one, look at what it drafts, and decide later whether it earns more trust.
Read moreAgents
Agents only send on their own if you turn on Auto for that agent. It is off by default, sends only when highly confident, to allowlisted recipients, inside a daily cap, and logs every send. One switch turns every agent back to Suggest. Agents have no tool to delete or forward mail, and you can switch AI off for a whole workspace.
AI screens every stranger who writes to you. Pitches go to a pile you review when you feel like it, and real people always get through.
Every “I'll send it Friday” is tracked, in both directions. You get a reminder before yours is due, and a polite nudge is drafted for theirs.
No more noreply. When someone replies to a receipt or a password reset, it lands in your team inbox threaded under the message that caused it.
Triage
Cold pitches, promotions and notifications are filed before you look. What matters lands in Primary with a one-line summary.
Drafts
For threads that need an answer, an agent pre-writes one in your tone from the thread itself. Review, tweak and send, or skip.
Follow-ups
Sent something and heard nothing? A polite nudge is drafted for you after the number of days you choose. You still press send.
Daily brief
What needs you, what is waiting on other people, and what your agents did overnight — as a card on Home and an email if you want one.
Ask your inbox
“What did the bank say about the payout?” gets an answer, with links to the exact messages it came from.
AI assistants
Connect Koltrix to Claude, ChatGPT, Cursor or any app that speaks MCP, the open standard for giving AI assistants tools. You sign in and approve it once; it sees only the mailboxes you can, and an admin can switch it off for the whole workspace.
Works with Claude, ChatGPT, Cursor and any other MCP client. See how the integration works or read the setup guide.
Search, read a whole thread, or ask what still needs a reply, in plain words.
Labels, archive, read and star. Everything it changes can be undone.
Replies wait in your Drafts. Sending is off until an admin allows it, and even then the assistant shows you the message and sends only after you say yes.
Claude and ChatGPT can read, search, organise and draft. Sending is off by default: an admin can turn it on, and the assistant must show you the message and get your yes before it sends.
Coming next
Listed so you can decide with the real picture. None of these is available today, and none of them is sold on any plan.