Sabitlenmiş Tweet
Introducing a library to make tailwind DX so much better...
typewind ✨
The magic of Tailwind combined with the safety of Typescript
github.com/mokshit06/type…
English
Mokshit Jain
647 posts
Mokshit Jain
@Mokshit06
19, introducing bugs @cognition • gap @caltech • i like compilers/agents/web/design • prev @replayio • built https://t.co/2Nnh053kmt & https://t.co/s84cCnKC9C
SF Katılım Haziran 2019
611 Takip Edilen1.9K Takipçiler
Mokshit Jain retweetledi
English
@sqs @cursor_ai Also, harness customizations introduce confounding factors: is it the same system prompt, tools, and skillset used for each model or different? (Of course, this is an issue with every existing model leaderboard.)
English
This is really cool. Seems like windsurf.com/leaderboard will be a real public good.
It's like @cursor_ai's parallel agents, but with the public leaderboard and blinding of model names.
Thinking:
- Who pays for the extra tokens?
- Testing 2 agents head-to-head requires truly isolated environments so they don't trample each other. Worktrees are only sufficient if you've set up your local dev setup right.
Windsurf@windsurf
Introducing Arena Mode in Windsurf: One prompt. Two models. Your vote. Benchmarks don't reflect real-world coding quality. The best model for you depends on your codebase and stack. So we made real-world coding the benchmark. Free for the next week. May the best model win.
English
Introducing Arena Mode in Windsurf: One prompt. Two models. Your vote.
Benchmarks don't reflect real-world coding quality. The best model for you depends on your codebase and stack. So we made real-world coding the benchmark.
Free for the next week. May the best model win.
English
Mokshit Jain retweetledi
Arena Mode is here and it's free for the next week (frontier models included).
Time to enter the Arena.
Windsurf@windsurf
Introducing Arena Mode in Windsurf: One prompt. Two models. Your vote. Benchmarks don't reflect real-world coding quality. The best model for you depends on your codebase and stack. So we made real-world coding the benchmark. Free for the next week. May the best model win.
English
feels very weird to say i’ve spent 0.5-1% of my life working on this…
but it’s finally out!
(and yes 5.2 is better than 4.5)
Windsurf@windsurf
Introducing Arena Mode in Windsurf: One prompt. Two models. Your vote. Benchmarks don't reflect real-world coding quality. The best model for you depends on your codebase and stack. So we made real-world coding the benchmark. Free for the next week. May the best model win.
English
i guess clawdbot is the new hype?
anyways, the set up is quite a process.
have devin set it up for u :)
Walden@walden_yan
Don’t waste ur time setting up @openclaw I had devin set it up for me. I didn’t even run a single command and now I’m talking to it on telegram tryclawdbot.com
English
Mokshit Jain retweetledi
"attention team, from now on, engineers are not allowed to review PRs. we will hire specialized software reviewers to do that instead"
Daksh Gupta@dakshgup
new blog post
English
Mokshit Jain retweetledi
wow, have been using @cognition's new PR review tool.
think whatever you want of devin. but this is incredible.
also makes it *really* obvious how vulnerable GitHub is.
English
Mokshit Jain retweetledi
honestly can't believe this is free lol, excited (ab)use this
Cognition@cognition
Devin Review is currently free and works on public or private GitHub PRs. You can use it in three ways: 1. devinreview.com 2. Swap github for devinreview in the PR url 3. npx devin-review {pr-link} - run this command inside the repo of the PR you want reviewed Check out the docs for more details: docs.devin.ai/work-with-devi…
English
what does the UX of reviewing code look like when agents are writing most of your code?
at some point you’re just looking at the code instead of understanding it.
try understanding it - devinreview.com
Cognition@cognition
Meet Devin Review: a reimagined interface for understanding complex PRs. Code review tools today don’t actually make it easier to read code. Devin Review builds your comprehension and helps you stop slop. Try without an account: devinreview.com More below 👇
English
@thdxr @samgoodwin89 increasingly convinced its 3. after seeing the same users use cli agents in the right panel of their editor
but yea alot of it feels posthoc
English
your pov is what i'd expect makes sense to most people but somehow it's not the case - i keep making threads asking people and there's not really a clear answer
ranges between
1. people don't like cursor agent
2. people like the mental seperation of agent vs editor
3. a big explanation might just be cost, people then posthoc justify it for other reasons
English
Maybe I'm a dumbass noob, but I can't function without all of this on one page. Cursor is still my main driver.
Git changes on left - click to see diffs.
Code is next, with diff overlays showing what the agent did.
Chat.
Agents list.
What am I missing about these CLI tools?
English
@aidenybai RSC but instead of remote code execution it’s local code execution
English
I’ve joined @Cognition to lead Design and build out their Design team. The Windsurf and Devin you know of today is about to get exponentially better.
Look forward to 2026 because we’re reshaping everything.
English
Mokshit Jain retweetledi
Congrats to the @antigravity team on the launch today!
fyi you missed a spot:
Varun Mohan@_mohansolo
Excited to launch Google Antigravity, our next generation agentic IDE, now powered by Gemini 3!
English
Mokshit Jain retweetledi
So Google just forked the Windsurf codebase and they even forgot to remove the Cascade branding in some places
Marvin von Hagen@marvinvonhagen
which ide did they use to build antigravity? windsurf? cursor?
English