So… this post was about me and @nate_vick and @selectiveci: “The intelligence layer on top of your existing CI”. It’s our product that does test selection, flake detection, queue distribution, and streams results to you or your agents in real time.
For those of you that don’t know, we started building Selective six years ago (yes, really), but never really launched it.
There are a bunch of reasons/excuses I could share but the main one (for me) was fear. Fear that it wasn’t good enough, or that it wouldn’t succeed.
Well, I’m happy to announce that we finally got our stuff together, put in the work, and we’re shipping it! We’ve been battle-testing it at ClickFunnels and it’s performing incredibly well (> 95% selection accuracy, > 70% compute savings).
We’d love for you to try it. We think it could really help your ballooning CI bill, while also saving you and your agents cycle time.
We’re eager to get feedback and figure out pricing. As an early adopter, we’ll take very good care of you. We’re even willing to white-glove implementation!
Hit me up via DM if you’re into it. 💚
Native integrations for CI build another layer of developer choice and control into the fastest Git forge on Earth.
Today, we’re excited to announce our partnership with @rwx , an agent-native CI platform now built directly into the Entire plugins. Try it out!
https://t.co/lNc9gMjEas
I’m excited to announce that @GoEntire now supports running CI through @RWX. Rebuilt from the foundation up for how agents work:
➡️ RWX: a graph-based execution model with automatic caching. Far less compute needed per loop.
➡️ Entire: Trails replace the pull request, keeping context, code, and verification in one place, even across repos and agent sessions.
https://t.co/Lo13Er8KEH
In this sharp AWS Developer Podcast moment, the conversation flips traditional coding wisdom on its head: when AI agents miss the mark, rewriting from scratch can be faster than fixing. A fun, forward-looking take on how developers should rethink prompts, context, and the true co
We talked with @dan_manges (@rwx co-founder, ex-Braintree CTO) about CI built for AI coding agents: smart caching, graph-based runs, agent sandboxes, and testing before you open a PR. #ruby#rails https://t.co/E4dGxcPv7z
@kushbhuwalka I can’t promise you 90% cost reduction, but RWX is very good. Content-based caching and graph-based execution eliminates so much unnecessary compute
@harjotsgill I’d be happy to give you a demo of RWX. Graph-based task execution and content-based caching result in the most efficient possible pipelines. Similar concepts as Dagger, but with a different approach.
You can get a lot of compute reduction from graph-based execution and content-based caching (Bazel or RWX) before you have to resort to running fewer tests. although that may be necessary eventually
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.
@thorstenball Do the CI checks run based on the definition on the branch, or on the definition based on main? I think that’s the key for making this safe. Can the agent change the CI pipeline?
@wgw_eth@depotdev@rwx@namespacelabs thanks so much for taking the time to share this! mostly fair assessment, and we're about a ship an improvement to the biggest source of overhead reported in here. great to see that rwx was still faster than some other alternatives.
@nateberkopec This isn’t true on AWS. They will SOMETIMES run on an older generation of compute. But they will sometimes run on the latest. They just don’t always have enough capacity in the most recent generation.
We've worked with the team at @cursor_ai to build and launch a first-class integration between Cursor Origin and @rwx
Origin’s landing page says, "Code is moving faster than any infrastructure was built to handle. Origin was designed for this moment."
Similarly, RWX is the CI platform uniquely built for this moment.