The After Effects Clock Returns, and Loops Grows Teeth

What people were saying

The day belonged to Loops. @linear announced that Loops can now trigger on project, initiative, cycle and issue changes, edit documents automatically, and send tailored Slack updates. The whole team turned up behind it: @artman called it a serious workflow engine and pointedly added that it's still early, @karrisaarinen framed the release as prioritizing product management use cases, and @jorilallo said they had effectively just hired a TPM, which is the kind of joke that is also a roadmap. @dizaytsev predicted "LoopOps" becomes a job title, which feels less like a forecast and more like a calendar invite.

The praise

A large share of the reaction was about the video, not the feature. @megxwayne noted the return of the After Effects clock of automation, and the design crowd piled in to say Linear's design team operates on a different plane. Elsewhere, people liked the small details more than the headline: continuing a completed run instead of starting a fresh conversation to fix one wrong assumption, and the general idea that agents get useful when the trigger is real work rather than another prompt box. One quote-poster argued that as software gets cheap to build, care and trust become the actual product, which is roughly Linear's entire pitch restated by a stranger.

The pushback

The sharpest questions were about provenance. Several replies under the announcement asked whether a document edit can be traced back to the project change that caused it, and whether a human can review a change to a shared spec before a Loop applies it. The worry is not that automation is wrong, it's that automation makes drift happen faster and quieter. Separately, @JorgeWritesCode floated the perennial "is this just Jira with a Notion theme" theory, and @notparbez needled the company over attributing a bug in native components to React Native.

Notable anecdotes

The day's most-engaged Linear mention wasn't from Linear at all. @championswimmer argued that the real bar for agentic infra is whether an engineer can pull out a phone, tell an agent to clone a repo, build, test, open a PR, read Slack and touch Linear, and have it actually work. He spent most of the post on the unglamorous part: pausing and unpausing VMs for tens of thousands of employees is expensive and hard to orchestrate. One reply put it bluntly: if your agent can't reach Linear and a real PR from a phone, it's a demo. And @brendanjshort relayed a GTM demo where a bot watches Linear for resolved customer bugs and tells the affected customers, which is a use case Linear's own customer requests feature keeps circling.