Buildkite

You deserve better CI. Cursor, OpenAI, Anthropic, Nvidia, Uber, Airbnb, Canva, Bazel, vLLM, Amp, Entire, Bun, Rails + more run their CI on Buildkite.

[Buildkite @buildkite]

The fastest step is the one you never need to run.

Caching is now generally available in Buildkite. Multiple techniques compounding together. The first of several new building blocks for faster, better CI.

[Lachlan Donald @lox]

We've shipped Buildkite Cache, the first of several building blocks we're putting in place to make CI radically faster.

We've learnt a lot from how the fastest companies in the world run their CI. Great caching isn't just one technique, it's several working together: what you reuse, where it lives, and which work you can avoid doing in the first place.

We're starting with a save and restore primitive to control cached files across jobs, on hosted or self-hosted agents. You choose what goes into the cache key, when to fall back to an older entry, and which builds can share the cache.

There's a lot more coming.

https://t.co/aegRNGwjSH

[Buildkite @buildkite]

Running an old Buildkite MCP server? Update yours to v1.28.0 so your agent can explain failed builds, retry jobs, and pause queues.

[Buildkite @buildkite]

The question isn’t how fast single builds run. It’s whether your CI gets faster or slower as your work compounds and multiplies.

More compute might buy you time... but it doesn’t change the answer.

[Lachlan Donald @lox]

We are either going to need more compute, or better CI.

AI driven dev means more code/tests and more frequent changes trying to merge and deploy to production.

E.g. 10× the changes × 5× the test suite cost = 50× the compute (and higher peak compute demand due to parallelism).

And that's before you rebase a few times to keep up with the fast moving trunk, then battle a merge queue due to flaky tests.

There are a few things everyone should be doing. First off, you make your test suites leaner and faster. High signal units, massive parallelism, thoughtful end-to-ends.

What needs to come next is compute reduction. You should only run the tests that relate to the change you are testing. The next step is getting even smarter. Models can be used to predict which subset of the related tests are most predictive of failure which lets you get to the failure fast and avoid unneeded subsequent test runs.

[✨ Phil Calçado @pcalcado]

All this talk about using only end-to-end tests for AI-written code makes me think the real winner of the AI race is going to be Buildkite.

[snoopy jpg @snoopy_dot_jpg]

@lox @CraigBarrDev i'm with this guy. just use buildkite, it's wonderful. and if you're still getting crushed by CI go fix your test and build pass

[Michael Hedgpeth @michaelhedgpeth]

Continue to love being on the @buildkite platform, as I see another GitHub Actions incident this morning. If you haven't thought of going in that direction, you should...

[Cynthia Toporowski @ItsCynnamoroll]

I love buildkite.

[Buildkite @buildkite]

Same workflow file, different agents. Run your existing GitHub Actions workflows on Buildkite's Mac fleet. No rewrite. Try now.

Read more: https://t.co/rZyYHw1749

[Buildkite @buildkite]

One for every build status you've ever felt personally.
Build Buddies, coming soon: https://t.co/Jl3ANlz8Fg

[Buildkite @buildkite]

Today, @spacexai released the Origin App Marketplace. The same CI setup used by @cursor_ai to build Origin is now available to you.

Our native Buildkite integration is included alongside key industry heavyweights such as @linear, @vercel and more.

Built by Cursor. Built on Buildkite.

[Cursor @cursor_ai]

Origin, Cursor’s code hosting platform, now supports over 20 integrations.

Connect your repos to tools for CI, deployments, code review, security, monitoring, and project management.

[Cursor @cursor_ai]

Origin, Cursor’s code hosting platform, now supports over 20 integrations.

Connect your repos to tools for CI, deployments, code review, security, monitoring, and project management.

[Nate Berkopec @nateberkopec]

@mattgapp Everything and everything but @buildkite is GOAT

[Brett @0xBrettj]

Mac orbs are in beta (https://t.co/g3bXlguSzo)

Our team has been iterating a lot on our Mac/iOS app primarily in orbs; I’ve often had an orb thread push the changes to my iPhone using a runner on my Mac, but @buildkite preflight and now mac orbs allow agents to test their changes on Mac in a remote environment

https://t.co/wUc4yfxZWK

[ClickHouse @ClickHouseDB]

@buildkite Watch Paul's full talk from Open House Australia 2026: https://t.co/wgOXhrUeIZ

[ClickHouse @ClickHouseDB]

Fast, flexible, economical. ⚙️

@buildkite ingests a trillion test executions a year. Paul Annesley on why they chose ClickHouse, and specifically ClickHouse Cloud, at Open House Australia 2026 👇

Start free with $300 in ClickHouse Cloud credits → https://t.co/vWC2a14W5x

[Buildkite @buildkite]

Great combo 💚

[Son Luong @sluongng]

Oh hey it's @bazelbuild 🤝 @buildbuddy 🤝 @buildkite

Stronk @Reddit !

[Buildkite @buildkite]

EXTRA! EXTRA! READ ALL ABOUT IT!

The Morning Build, Vol. 1 is out. A real, printed newspaper, hot off the press and on the streets of SF this week.

Didn't snag a copy? https://t.co/SI426FuD1O

#SFTechWeek

[Thorsten Ball @thorstenball]

@nikstep This was just in the orb using Buildkite with "bk preflight" to get an iOS build

[lauren @poteto]

i just automated our release and QA process!
using Grok Bot i made two Team Bots: sandcastle, my release manager, and poteto, my engineer bot. i added them to Slack and made a new channel for managing releases.

first i tell sandcastle that i want to do a cut. it'll DM all the contributors with links to their PRs so they know what goes in the next release and can DM the bot back if they have any blockers or objections.

sandcastle then kicks off the build, watches it automatically, and starts up an automated fuzz swarm. usually this is 10+ agents running on Grok 4.7 xhigh to run the build, click around and use it like a real user (using our verification skills and feature map), some are directed while a few are "chaos monkey" clicking things randomly to see what breaks.

whenever it finds an issue, i tell it to @ mention my engineer bot poteto, to spin up a Cursor Project to own triaging and fixing all the high pri issues. depending on the severity i may ask it to cherry pick the fix into our release branch and cut a patch release, or if it's a pre-existing issue i still fix it but include in the next release instead

that's one more tedious task that Grok Bot now handles for our team

[Cynthia Toporowski @ItsCynnamoroll]

I just realized I can solve all my problems by moving CI from GitHub actions using Namespace actions, over to buildkite.

Namespace devboxes are still essential to my workflow so all agents have mac machines to compile and test on. This change would only be so I stop hitting the stupid GitHub actions limit when I'm not even using their compute.