We're adding support for AGENTS.md to Claude Code.
Starting today in version 2.1.277, if there is no CLAUDE.md in a folder, Claude will check for and use AGENTS.md.
You can toggle this behavior in /config.
Proud of this moment, moment of a lifetime.
Vulcan, built by team @Razorpay, in India, shown to the Honourable Prime Minister of India, Shri Narendra Modi today.
A foundational model for payments, built in India, for the world. 🇮🇳
A special moment for @Razorpay today. 🇮🇳
Our CPO @khilanharia presented Vulcan - Razorpay’s foundation model for payments to Hon’ble Prime Minister Shri Narendra Modi today.
A proud day for everyone who built it. And we’re just getting started.
Built in India. 🇮🇳
A lot more to come! 🔥
This was an interesting meeting for me. The amount of potential in Indian space startup scene is impressive. All I can say is that I found sincerity from the top-most people in India for the application of space for the common people. Inspiring to see that!
#NSpD2026
4 billion payments. 3 trillion data points. One model trained on all of it.
Meet @Razorpay Vulcan - India’s first transformer-based AI foundation model for payments.
Built, trained and hosted in India, in partnership with @nvidia and @awscloud.
Until now, every payments problem was solved separately: routing, fraud, risk, personalisation and more.
We asked: What if one model could understand how money moves?
And like LLMs are trained on text to understand language, Vulcan is trained on payments to understand how money moves.
Already running in beta across 51,000+ businesses, Vulcan is delivering:
- 8–10% improvement in payment success rates - the ultimate measure of whether a payment simply works.
- 8x more international card fraud detected.
- 5x more fraudulent or disputed transactions identified.
- 1–2 lakh more purchases completed every month through better checkout personalisation.
And we’re just getting started. The best part? Every payment Vulcan sees makes the next one smarter.
We’ve spent years building the infrastructure that moves money for India. Now, we’re building the intelligence that understands and improves it. 🔥
Ordered from @SwiggyInstamart today and the entire order arrived reeking of dung.
Support just repeated scripted lines. This isn't about a refund anymore, it's a public health issue if the store is contaminated.
@SwiggyCares@fssaiindia@dineshgrao please inspect this store.
Me: I need better work-life balance
Also me: setting up a VPS with Claude Code after seeing @levelsio's guide
Now I can code from the gym. And the bathroom.
We're not okay.
I have a sneaky suspicion that
a) Gemini 3.0 is going to be a near AGI moment
b) It’s going to be released sometime before end of year, maybe before thanksgiving
c) they previewed it to Tim Cook and Apple has licensed it, in return for training data to Google
I am under NDA at a bunch of places but I studiously try to know nothing, so none of this is inside information in any way.
I’ve just been reviewing the news flow for the last month and the last missing piece fell into place today.
This means that the table resets soon, making Siri a serious competitor to ChatGPT.
Apple still has legacy UI, apps, they don’t own the model, etc. My guess is that they will take 3 years to make a AI native device (it generates pixels on demand, no code)
OpenAI potentially has an AI native device out next year, has an app store coming soon.
Google is the only one who controls device to model at this point.
xAI, gosh, Elon has to make a phone. You can’t wait around for 20k bots or 50k cars to get you distribution.
Anthropic is not playing for consumer.
Meta is potentially screwed if there is a big shift in consumer prefs next year. Remember Shopify is a FB/Insta feeder company and they signed a deal with OpenAI.