What building blocks should Meteor ship out of the box?

Hi everyone :waving_hand:

This thread went quiet in April and I think it deserves a second round. A lot has moved since then, some needs have probably changed, and most of us now write a good part of our apps with an AI assistant sitting next to us, which changes what “a first-party package for X” actually costs. So here is a short catch-up, two polls, and a question for the people who posted above.

What changed since March

  • Meteor 3.5 shipped with Change Streams as the default reactivity engine, 3.5.3 is the current stable, and 3.6 is at rc.1 with meteor create --pwa (#14473). Capacitor (#14633) and OpenTelemetry (#14086) are on the 3.7 milestone, so two items of the original list (mobile and observability) have a path now.
  • @mvogt22’s three core PRs are still open: mongo-schema (#14279), thread-context (#14281, 3.7 milestone) and the job queue (#14286). Jobs and schemas were the two items everyone agreed on above, and they are still waiting :face_with_diagonal_mouth: Please review it if you have time ! :partying_face:
  • The official roadmap (Roadmap | Docs, last updated April 30) lists Capacitor, release CI, OpenTelemetry, TypeScript and the test toolkit as the next priorities. None of the building blocks discussed here (jobs, uploads, email, caching, i18n, admin, migrations, rate limiting) is on it yet, and the roadmap itself says feature requests belong on the forum. So this thread is the right place to build that signal :grin:
  • Meteor now has a governance document, a security process and a contributor sponsorship program, so “who could take this” has more answers than it had in March.
  • meteor/agent-skills exists, with 17 Meteor-maintained skills, so an assistant now writes decent Meteor code out of the box. Which brings me to the next point.

The AI angle

Honest take: with an assistant, a first-party package for most items above is days of work for one person, not weeks. Writing it is no longer the expensive part. The expensive parts are deciding what Meteor should own, reviewing it, and maintaining it for years, and the core team already has a full plate. So the useful signal from this thread is not “who could build it” but “what do people actually need and use in 2026”, so contributors pick the right things and the core team can say yes or no quickly.

Two polls (30 seconds)

Which building blocks would you want first-party first?
  • Background jobs and scheduling
  • Schema and validation
  • File uploads and storage (S3, R2, local)
  • Email: templates, queue, providers
  • HTTP API conventions and rate limiting
  • Caching
  • Logging and observability hooks
  • Admin panel / auto CRUD
  • i18n
  • Push notifications
  • Full-text search
  • DB migrations
  • Clearer docs on what is core, official, recommended or community
  • Something else (tell me below)
0 voters
When you needed one of these in 2026, what did you do?
  • Used a community package
  • Wrote it myself (with or without AI)
  • Used an external service
  • Did without
0 voters

Your turn

  • If you posted above (@mvogt22 @msavin @harry73 @jam @paulishca @victor @fredmaiaarantes): is your take still current? One line is enough.
  • If you missed this thread in March: what did you have to build or plug in on your last Meteor app? One line, or a link to the package you used.
  • If you want to pick one of these up, say which one here so we don’t build the same thing twice. Community package first, docs page, core later if it fits.

Cc @italojs @nachocodoner @grubba, no pressure, just so the signal reaches you when you plan 3.7 and beyond :folded_hands:

2 Likes