Aashish Peepra

I won $500 💰 last night from @simverse_ How? She bet I couldn’t build @web3meets in 1 days. Guess what I did 🚀

Had fun meeting @CommendaHQ customers, the coffee wasn’t that great

Live updates at @ycombinator Bangalore startup school, it’s too hot here.

Don’t go clubbing one night before an event like me and @shrey_of_light. Be smarter!

The AI does the work while you learn how to play on the weekends

Computer use is now in Claude Code. Claude can open your apps, click through your UI, and test what it built, right from the CLI. Now in research preview on Pro and Max plans.

@probablyaayushi Woaaaahhhhh. I’ve a nice YouTube feed.

@peeps_ap pretty obvious 🙈 you forgot logging off.

Subletting my place for 2 months, ₹25k/month, all inclusive (rent + utilities, no headaches). Super calm, cozy spot in Koramangala 1st Block. Walkable to Cult, Starbucks, Third Wave & Wipro Park. If you or someone you know needs a short-term stay, DM me.

Excited to announce our partnership with @noicedotso. Noice traders can now incorporate a Delaware C-Corp directly through the app, powered by Commenda. First embedded incorporation of its kind, and just our first use case. We are working toward making all of Commenda's services available through MCP, so any affiliate app or agent can run incorporation and compliance workflows without sending users elsewhere.

It was fun building Incorporation as API for @noicedotso Working on bringing global commerce MCPs to Claude built on top of @CommendaHQ

you can now setup your delaware c-corp entity with trading fees. manage all your docs, entity details & application status straight from noice. feature unlocks above $1m market cap. in partnership with @CommendaHQ.

@LowKeyDevs I’m seeing 16 extra PRs next week

First tweet from my 13-inch MacBook Air (M4) — 10-core CPU, 10-core GPU, 16-core Neural Engine, 16 GB RAM, 512 GB storage, 35 W charger Mueheheheheheh!!!!!

Higher chances to get a job at a startup. Apply at @CommendaHQ

One general problem with LLMs is the missing perception while generating answers. Have been reading a lot about arts and niche creative skill generation using LLMs and why they fail. You can always fine tune them using SFTs, LoRA and so on. But that will require a normal human to train their own models (maybe once GPUs are cheaper we can afford this). But for now, I’m experimenting with adding a perception layer on top of an existing LLM model (another reason why you don’t wanna train a model is because then you’ll be competing against tech giants for latest model, which is the wrong race). This should make your answers more tailored towards your needs on the run time. Will share updates on weekend!

@LowKeyDevs You need this for the oncalls

Anyone need?

@bcherny Check the section "We made repository knowledge the system of record" from openai.com/index/harness-…

3. Invest in your CLAUDE.md. After every correction, end with: "Update your CLAUDE.md so you don't make that mistake again." Claude is eerily good at writing rules for itself. Ruthlessly edit your CLAUDE.md over time. Keep iterating until Claude's mistake rate measurably drops. One engineer tells Claude to maintain a notes directory for every task/project, updated after every PR. They then point CLAUDE.md at it.

I'm Boris and I created Claude Code. I wanted to quickly share a few tips for using Claude Code, sourced directly from the Claude Code team. The way the team uses Claude is different than how I use it. Remember: there is no one right way to use Claude Code -- everyones' setup is different. You should experiment to see what works for you!

@thdxr I don't get this tweet. What was the outcome?

everyone's talking about their teams like they were at the peak of efficiency and bottlenecked by ability to produce code here's what things actually look like - your org rarely has good ideas. ideas being expensive to implement was actually helping - majority of workers have no reason to be super motivated, they want to do their 9-5 and get back to their life - they're not using AI to be 10x more effective they're using it to churn out their tasks with less energy spend - the 2 people on your team that actually tried are now flattened by the slop code everyone is producing, they will quit soon - even when you produce work faster you're still bottlenecked by bureaucracy and the dozen other realities of shipping something real - your CFO is like what do you mean each engineer now costs $2000 extra per month in LLM bills

This system for information generated by the user post the system was installed is easier, gathering the information from different sources and then making sure it’s always up to date is hard. Why not fetch and use the information at run time and not become the source of truth for things you don’t need to be?

@LowKeyDevs I know what this is.

sneak peek of my new build 👀

revamped my portfolio website (lowkeydev.me)