Tomas Reimers

1.1K posts

Tomas Reimers banner
Tomas Reimers

Tomas Reimers

@TomasReimers

Cofounder / CPO @withgraphite; prev. Eng @facebook NY.

New York, NY Katılım Kasım 2012
1.1K Takip Edilen1.5K Takipçiler
Sabitlenmiş Tweet
Tomas Reimers
Tomas Reimers@TomasReimers·
AI means that more code will be written. But that also means… More code will need to be reviewed. More code will need to be tested. More code will need to be merged. More code will need to be deployed. @withgraphite solves those needs.
English
3
3
34
6.5K
Max Musing
Max Musing@MaxMusing·
man, the new @linear review feature is actually good. I can finally work with 10k+ line diffs. still laggy in a few places but miles ahead of GitHub and Graphite.
Max Musing tweet media
English
14
2
324
27.4K
nadav
nadav@NadavAHollander·
begging the @github @graphite overlords -- please let me pass in some form of "review map" into a PR that breaks up the diffs into annotated sections a la devinreview.com and other tools. makes reading agent-generated code 10000x easier
English
2
0
1
568
Brandon McConnell
Brandon McConnell@branmcconnell·
at this point, most project management solutions have great integrations with git services/providers like github but who is building the project management solution that is perfectly blended with git? @graphite is close, but still detached from this flow
English
3
0
2
484
Rahul Jain
Rahul Jain@rahulj51·
Afaik @graphite is the only review tool that allows you to chat with a PR. You could, of course, do that in a claude session too (give it the link to a PR) but Graphite's UX is better and more tailored to reviews. I wish their chat interface was a full fledged Agent that allowed posting comments or performing actions on the PRs.
English
4
0
5
605
nadav
nadav@NadavAHollander·
is there a code review tool a la graphite that intelligently groups file changes into a readable sequential narrative? @cognition's devinreview.com does this nicely but is very slow - blocks on soliciting the full agentic review before rendering diffs in intelligible order
English
4
0
5
555
Tomas Reimers
Tomas Reimers@TomasReimers·
Excited to talk at the @mlopscommunity conference in March! Come listen about agents and the future of review (link below👇)
Tomas Reimers tweet media
English
2
1
9
678
edwin
edwin@edwinarbus·
The newly expanded Cursor campus now includes a former event space, a loft, an office building, an old movie theater, the 8 bus stop, and the USPS North Beach Carrier Annex.
edwin tweet media
English
23
6
269
21.3K
Sam Lambert
Sam Lambert@samlambert·
sometimes you just go into a PlanetScale Slack channel and see people just ripping PRs and unblocking each other
Sam Lambert tweet media
English
8
0
99
23.4K
n2parko
n2parko@n2parko·
today we introduced Cursor Blame it turns out it's useful to persist the "why" behind your code so your teammates can understand code lineage ...and future agents can use this context to make better decisions
English
57
44
1.2K
139.5K
Andy Coenen
Andy Coenen@_coenen·
I wanted to share something I built over the last few weeks: isometric.nyc is a massive isometric pixel art map of NYC, built with nano banana and coding agents. I didn't write a single line of code.
Andy Coenen tweet media
English
355
655
7.9K
1.5M
Jonathan
Jonathan@gravyskyy·
@TomasReimers @graphite @cursor_ai Congrats! This makes a ton of sense. If AI increases code output, review throughput becomes the limiter. Curious what you’re most excited to ship together: smarter diffs, auto-stacked PR flows, or agent-assisted review that actually drives changes?
English
1
0
0
30
Tomas Reimers retweetledi
Ben Davis
Ben Davis@davis7·
full triage and fixes submitted for all open issues on btca done in a proper graphite stack just told it to do this, came back 10 mins later and it's done. feels like the ultimate async model...
Ben Davis tweet media
English
6
1
76
6.1K
Tomas Reimers
Tomas Reimers@TomasReimers·
@maxmarchione Have you considered rebranding them as plasmids? Or renaming to Fontaine Futuristics?
English
1
0
6
2.9K
Ayush Sood
Ayush Sood@ayushsood·
@ankrgyl @rauchg @TomasReimers said the same thing to me but more generalized - you have 3-4 “cards”. Ie you want to tech innovate (research)…that’s a card, solo-found…card etc. by themselves not bad, but you can only have 3-4
English
1
0
1
863
Ankur Goyal
Ankur Goyal@ankrgyl·
One of the most important questions as a startup is when to go deep on solving a problem in a truly innovative way vs. just enough to check a box. @rauchg has a nice framework for this — something like every startup gets "5 technical bets".
English
9
2
122
26.1K

Keşfet

@andrewqu @peer_rich @MaxMusing @linear @NadavAHollander @github @graphite @branmcconnell