Rohan Varma
@rohanvarma
Followers
8K
Following
3K
Media
68
Statuses
718
@cursor_ai @adoptclarity, @tryexplo, @czi, @palantirtech
San Francisco, CA
Joined March 2009
The fact I think this is a bad idea, probably means it’s worth trying. @austinnickpiel and I spent one year trying to build this. Maybe the model capabilities and agents weren’t good enough then. I’ve also chatted with maybe ~10 startups (many in YC) that are trying to do this
Cursor for Product Managers @amiklas AI is great at writing code, but product success depends on deciding what to build. We think there’s an opportunity for an AI-native system focused on helping teams figure out what to build, not just how to build it.
16
2
56
The future of development lives in Agent Development Environments (ADEs). The core jobs IDEs were built for no longer reflect how frontier engineers spend most of their time. Agentic coding introduces a new set of requirements that ADEs are built around: - spawning multiple
Introducing the Codex app—a powerful command center for building with agents. Now available on macOS. https://t.co/HW05s2C9Nr
12
5
99
Carpe diem 🙂
2
0
24
I'm hacking this weekend on some AI-native code review prototypes. In 1 hour and 10 prompts, with the Cursor CLI, I went from an empty repository to a fully functioning code review surface that integrates with Github, has a performant diff viewer, handles stacked PRs, and even
13
1
89
We are steadily automating product building at Cursor. We started with code generation and now we’re heading in both directions. Goal setting <- 👀 Planning <- Ask Mode, Cloud Agents Task tracking <- Agent + MCP Architecture design <- Plan Mode Code generation <- Cursor Agent
11
4
137
We're proposing an open standard for tracing agent conversations to the code they generate. It's interoperable with any coding agent or interface. https://t.co/jO4DIoIl6A
agent-trace.dev
A standard format for tracking AI-generated code.
110
209
2K
How many collective engineering hours are spent every day trying to return to the GitHub PR page after clicking a link navigates you away? Default open links in a new tab. @github 🙏🏾
15
0
76
My top Cursor CLI workflow right now: > git clone <new_repo_url> > agent > "Install dependencies and run the local dev server" I can come back in 5 minutes and have a running dev server - no more fudding around in README files
3
3
51
I love the Cursor CLI for coding, but the precision of seeing a CI error, going to Cursor, and clicking "Fix in chat" is so nice
5
1
67
Just moved desks for the 3rd time in 3 weeks across different buildings Exciting to see our team grow @cursor_ai 🚀🚀🚀
11
3
210
If you liked single agents with browser access, you'll love this 👀
2
1
61
Reply with your startup, product, or tool ideas 👀 @n2parko and I are doing a mini-hackathon! We'll take your ideas from ChatPRD -> Cursor Cloud Agents -> Vercel and send a link to the product in response!
Two things live this week at @chatprd for two types of PMs: PMs who do PRs: chatPRD > @cursor_ai integration PMs who don’t: document versioning All the goodies in our weekly release notes: https://t.co/BA1O3qIlIt
17
3
41
🤌🏾 teams cooking on the CLI
6
2
62
Codebase scanning for bugs is a harder problem than it might seem. Stay tuned for more on this from Cursor and Bugbot 👀
I took Opus 4.5 Max Mode for the ultimate test. I asked it to search my code base for bugs I took a lot of pride in this code, all hand rolled, by the sweat of my brow. It was mine and it was good. Claude found 15 bugs I was devastated So I went to the first bug, and asked
6
1
106
Code review is the real bottleneck. AI helps you write more code, but if reviews don’t scale, velocity doesn’t either. Cursor Bugbot with Autofix unblocks teams: instant reviews, fewer back-and-forths, and fewer bugs shipped.
29
14
302
Apple loves shipping iPhone features that trigger unexpectedly: - Siri - The entire U2 album - AI Mode
3
0
28
This work by @cursor_ai is, I think, the coolest AI breakthrough since GPT-4. (And there are plenty of candidates!) https://t.co/btmcIK3rlY
simonwillison.net
Wilson Lin at Cursor has been doing some experiments to see how far you can push a large fleet of "autonomous" coding agents: This post describes what we've learned from …
68
233
3K