Endlessly refactoring and rewriting code at @Cloudflare | Ex-Founding Engineer at @dyte_io (@ycombinator W'21) 💻

Joined June 2020
164 Photos and videos
Pinned Tweet
An AI agent can write your code in minutes. But someone still has to review, merge, deploy, and monitor it. What if the agent could do that too? Feature flags are the missing piece. They let an agent ship code behind a flag, test it on real traffic, ramp the rollout, and kill it instantly if things break. No human in the loop until you choose to be. Today we're shipping Flagship to make this possible - feature flags native to @Cloudflare's network, OpenFeature standard. Move fast, break nothing. blog.cloudflare.com/flagship
10
30
156
32,798
Rohan Mukherjee retweeted
Bengaluru, thank you for showing up today. Met the coolest people, building the coolest things on the @CloudflareDev platform. A big shoutout to @cengines_studio for hosting us.
3
10
81
6,233
We just wrapped up our @CloudflareDev buildathon and it was a full-house! Thank you for showing up and shipping with us! We should do this more often, shouldn't we? 👀 Shoutout to @cengines_studio for letting us use their beautiful office
2
4
51
3,118
Flagship now has API support. You can now interact with @CloudflareDev Flagship programmatically - enabling workflows, automations, integrations, and agent-driven usage beyond the UI. 🧵
Flagship API reference is now available. Manage apps and feature flags entirely via the API without touching the dashboard. developers.cloudflare.com/ch…
2
3
22
1,912
For AI workflows, Flagship is also available as part of the @CloudflareDev skill This gives agents the context they need to understand and work with Flagship capabilities alongside the broader Cloudflare platform. github.com/cloudflare/skills
1
2
155
Whether you’re creating developer workflows or agent-powered experiences, the pieces are now available to start building. Feedback welcome! Excited to see what y'all build with this! 🚀 developers.cloudflare.com/fl…
64
Rohan Mukherjee retweeted
Come Build with Cloudflare devs this Weekend! Met with @fayazara @roerohan and @11sh1ta this week and talked about the developer ecosystem in Bangalore. Within 15 minutes (efficient af meeting) we concluded that we gotta do more build sessions with the community!
2
4
21
1,276
Peak slop UX from ChatGPT: "Try Plus free for 30 days" with buttons for everything except trying Plus free for 30 days 😅
6
16
921
update on the Fable 5 one-shot refactor attempt: closer than any model has gotten, but it still missed some bits... guess we wait for the next one.
using Fable 5 on @opencode tonight by feeding it the refactor I’ve been avoiding for months. I'm trying to one-shot it, and the repo has no tests. this is the real eval
2
254
Rohan Mukherjee retweeted
we're hosting the good peeps from @CloudflareDev at ours this saturday for a good ol' fashioned build sesh. y'all don't wanna miss this one. apply now! ⬇️
Bengaluru - @CloudflareDev is hosting a builders-focused hackathon this saturday at the @cengines_studio office in Indiranagar. Sign up today. luma.com/vsyebdod?build=true
3
3
18
1,815
We're hosting a long-awaited @CloudflareDev meetup this Saturday, 13 June, at @cengines_studio! If you're in Bengaluru this weekend, come hang out with us, and ship something cool!
Bengaluru - @CloudflareDev is hosting a builders-focused hackathon this saturday at the @cengines_studio office in Indiranagar. Sign up today. luma.com/vsyebdod?build=true
1
3
14
689
Jevons' Paradox is real. The lower the cost of coding, the higher the demand for software.
This is a super exciting release - Claude Fable 5 is the same underlying model as Mythos but with added safeguards. The benchmarks are great and it's SOTA on everything by a margin but I'll add that *qualitatively* also, this is a major-version-bump-deserving step change forward (imo of the same order as Claude 4.5 was in November), peaking especially for long problem-solving sessions on very difficult problems. You can give it a lot more ambitious tasks than what you're used to, the model "gets it" and it will just go, and it's never felt this tempting to stop looking at the code at all (but don't do this in prod!). The model still has quirks that people will run into and the safeguards are configured to be a little too trigger happy for launch, which can hopefully be tuned over time. I feel a lot of things changing as working software increasingly comes out on a tap. The Jevon's paradox kicks in and I feel my own demand for software growing substantially. You can ask for anything - explainers, visualizers, dashboards, bespoke single-use apps (e.g. a full wandb that is hyper-specific just for your project), you can 10X your test suite, auto-optimize code, run giant research projects with custom HTML for the results, anything! "Free your mind" (Matrix ref). Really looking forward to all the things people build!
1
2
426
using Fable 5 on @opencode tonight by feeding it the refactor I’ve been avoiding for months. I'm trying to one-shot it, and the repo has no tests. this is the real eval
1
16
645
manually debugged something today that the agent couldn’t figure out. maybe we actually get to keep our jobs for a little longer?
3
7
287
everyone's so locked into WWDC discourse that nobody noticed @janja_garnbre just sent Bibliographie (9b )
2
9
3,171