Who
I wanted company email on day 1 of an idea without paying Google or Microsoft per seat while the product was still a maybe.
For founders who think mail.config.ts in git should be law, ahead of another admin dashboard.
What
I built codemail, private https://github.com/michaelmonetized/codemail (selection label HurleyUS/codemail; origin is michaelmonetized). HEAD e9fece4. 27 commits. Default main. Version 0.1.0.
Thesis: email infrastructure for founders: config as code, per-domain pricing, unlimited mailboxes. Adjacent to Gmail, not a seat-for-seat swap.
Shipped (weekend MVP arc): Turborepo monorepo; @codemail/config (Zod + defineMailConfig); @codemail/cli (setup/deploy/status/dns/logs/users); @codemail/smtp on Fly; Convex backend; Next 15 web mail + dashboard; Clerk; Resend outbound; marketing Free Forever / Simple $8 / Managed $80 / Self-Hosted.
Gaps: PLAN still marks IMAP out while README + Simple tier market IMAP; GitHub public probe 404 (private) despite OSS CTA; codemail.vercel.app MIDDLEWARE 500 vs codemail-web.vercel.app 200; root mail.config.ts is informal t3.chat demo shape; codemail.dev unchecked.
Where
Code: https://github.com/michaelmonetized/codemail (private).
Live marketing/app: https://codemail-web.vercel.app
SMTP: codemail-smtp.fly.dev (per TODO)
When
2026-02-12. SMTP, then web mail, then dashboard, then Convex, then API, then auth/SEO, then Fly SMTP fixes, then marketing rewrite to business-plan thesis (same-day blast).
2026-02-13. public marketing routes + (private) auth route group.
2026-02-15. 62fcaed better sales positioning (#1).
2026-02-18. e9fece4 design updates. HEAD.
Why
Day-1 company email should be a git push, not a $600/year seat tax. Config in the repo beats another dashboard click-path when the product is still a maybe.
Engagement Q: Ship IMAP before opening the repo public, or cut IMAP from the landing first?