Kashyap Murali
@katchu11
Followers
101
Following
4K
Media
28
Statuses
112
a lil bit of this, a lil bit of that eng. MTS @ Anthropic
San Francisco Bay Area
Joined October 2014
excited to hear how y'all like the explore subagent!
Haiku 4.5 is a workhorse that makes the coding experience in Claude Code feel really fast. While Sonnet 4.5 remains the default, Haiku 4.5 now powers the Explore subagent which can rapidly gather context on your codebase to build apps even faster.
7
2
103
🧵 Haiku 4.5 🧵 Looking at the alignment evidence, Haiku is similar to Sonnet: Very safe, though often eval-aware. I think the most interesting alignment content in the system card is about reasoning faithfulness…
2
11
70
Anthropic CEO Dario Amodei met today with Prime Minister @narendramodi. We're looking forward to growing our Indian team and supporting India's AI ecosystem as it develops the next generation of dynamic companies.
Today I met with PM @narendramodi to discuss Anthropic's expansion to India—where Claude Code use is up 5× since June. How India deploys AI across critical sectors like education, healthcare, and agriculture for over a billion people will be essential in shaping the future of AI.
128
367
5K
Extremely excited about the new Claude Code features and hacking on them with 4.5 sonnet! The rate of progress has been staggering and the new UX + features make it all the more easier to use. Back to hacking on what’s next :)
Introducing Claude Sonnet 4.5—the best coding model in the world. It's the strongest model for building complex agents. It's the best model at using computers. And it shows substantial gains on tests of reasoning and math.
0
0
1
We're sorry - and we'll do better. We're working hard on making sure we never miss these kind of regressions and rebuilding our trust with you.
We've published a detailed postmortem on three infrastructure bugs that affected Claude between August and early September. In the post, we explain what happened, why it took time to fix, and what we’re changing:
176
39
2K
really proud of the entire team for chasing this down and being transparent about it Claude Code is operating at a scale never seen before and that means bugs on an entirely new scale if you experienced any degradations before I'd suggest you try it again, should be cruising
In our investigation, we uncovered three separate bugs. They were partly overlapping, making diagnosis even trickier. We've now resolved all three bugs and written a technical report on what happened, which you can find here:
41
18
600
We just shipped a UX update for Claude Code: you have more control over when extended thinking kicks in using '/t'
62
64
848
Claude Code's GitHub integration is now generally available with a simplified API, ready-to-deploy templates, and support for more GitHub events beyond @-claude mentions.
60
239
3K
Ridiculous that OpenAI claimed 74.9% on SWE-Bench just to prove they were above Opus 4.1’s 74.5%… By running it on 477 problems instead of the full 500. Their system card only says 74% too.
151
262
5K