Launching today
CatchTheMail
Forward an email, get its priority and a ready reply
2 followers
Forward an email, get its priority and a ready reply
2 followers
Every AI email tool wants full mailbox access before it proves anything. CatchTheMail wants one email. Forward it to your private address, or set one rule that forwards a category automatically. Seconds later it comes back with a priority, a category, a short summary and a draft reply. There is no OAuth screen and no mailbox integration, so you grant access one email at a time and revoke it by deleting one rule. 7 day trial, 50 analyses, no card.
Hi Product Hunt,
I built CatchTheMail after missing a client email for three days. It was
sitting between forty newsletters, same font, same size, and by the time I
opened it the job was gone.
Every tool I looked at wanted full access to my mailbox first. Read, write,
delete, everything, before it had proven it was worth anything. That felt
backwards to me, so I built the opposite.
How it works:
1. You sign up and get a private forwarding address.
2. You forward any email to it, or set up a rule that forwards automatically.
3. Within seconds you get a reply with the priority, the category, a summary
and a draft response.
4. If you want to answer, you can edit the draft and send it to the original
sender straight from the dashboard.
There is no OAuth screen and no inbox connection, because the product never
sees your mailbox. It only sees the emails you choose to forward. That is a
deliberate architectural decision, not a temporary limitation, and it will
not change.
What is running underneath: Next.js on Vercel, Supabase for data and auth,
Postmark for inbound and outbound mail, Claude Haiku for the analysis.
Payments go through Lemon Squeezy as merchant of record.
Pricing: 7 day trial with 50 analyses and no card. After that $16.99 a month
for 100 analyses, $27.99 for 600, $46.99 for 2000. Yearly is ten months for
twelve.
I am the only person on this, and today is day one, so there are no user
numbers to show you. What I can tell you is that the whole pipeline has been
running on my own inbox for weeks, and the last two bugs I fixed came from
reading my own dashboard, not from tests.
I would genuinely like to hear where this breaks for you. If the analysis
gets something wrong on your email, tell me and I will look at the actual
prompt behaviour behind it.