Fetching from the wire…
01
02
03
04
05
06
07
08
Public story · 2026-07-30 · high
Cursor's July 29 iPad release adds a PR-review inbox with Apple Pencil markup, a day after the company's Start plan launch.
Why now: Covered in the July 30 briefing, one day after the July 29 release and two days after the July 28 Start plan.
Cursor shipped an iPad app on July 29 that skips the one thing you'd expect from a tablet editor: editing.
That matters for anyone running several coding agents at once. Review and approval no longer need a desk; a PR can get an okay from a phone or a couch.
The app centers on an inbox for pending work and PRs under review. Sidebar chats stay pinned, so you can track multiple agents in parallel rather than tabbing between chat windows.
PR review is full-featured: comments, checks, and approvals, all in the same surface. Apple Pencil support is for marking up screenshots, not writing functions.
Multi-PR sessions now work. Previously, only the last PR from a chat was reachable.
Cursor also added Bitbucket and Azure DevOps SCM support, plus in-app team switching.
All of it ships on paid plans, starting July 29.
The timing stands out. This is Cursor's second announcement in two days, right after the July 28 Start plan. Two releases in 48 hours reads like a company committing to a bet fast.
Nobody opens an iPad to write code. Plenty of people want to check on four agents from the couch and leave a comment with a finger-drawn circle around the bug.
Cursor built for that second group. Coding tools now compete on who owns the review queue, not on autocomplete quality. That shift happens once agents generate more code than one person can read line by line.
Watch whether competitors ship a comparable mobile review surface in the months ahead, or whether this stays a Cursor-only bet.
Each link below shares sources, entities, or timing with this story.
Within five days of each other, both Claude Code (v2.1.158, May 31) and Cursor (3.6, May 29) shipped remarkably similar architectures for autonomous agent execution. Both use a classifier subagent that reviews each pending action against conversation context and decides: allow...
Slack Code shipped August 20. You tag a coding agent in a conversation, it opens a project-specific channel, and the whole team watches the work happen: diffs, live HTML previews, feedback the agent picks up mid-run, approval gates before anything ships. Each channel keeps an...
Here's the experiment: a team of cooperating agents rebuilds SQLite in Rust from scratch, using only the 835-page manual. No source code. No test suites. No internet. Then it has to pass a held-out sqllogictest suite. It worked. Cursor published the research (Wilson Lin, July...
The largest available cost lever might not be the model. It might be a prompt telling the model to calm down. DietrichGebert/ponytail is a portable agent skill with a simple thesis: force coding agents toward the minimal solution. YAGNI first. Standard library before custom co...
Gergely Orosz published the first serious look at what AI coding actually costs at scale, and the numbers are wild. The Pragmatic Engineer covers "tokenmaxxing," a trend where engineers compete on AI token consumption leaderboards. At Meta, one engineer averaged 281 billion to...
Cursor's 2026-08-19 changelog adds Subscriptions to cloud agents, and it changes what an agent *is* in a way that a normal feature bullet undersells. An agent can now subscribe to an event stream and act on it without you prompting anything. Agents automatically subscribe to P...
MindPattern daily
One email a day at 7 AM. Sources and a take on every story. Unsubscribe anytime.