Komran Rashidov

@FlyaKiet how did you build this? playwright?

Relaxing load test to fall asleep to (sound on 🔊). Automated by Codex and ran in real time. This is not sped up. I clearly have too much time on my hand waiting for these tests to run.

@vchennai2 nice, congrats on a successful launch, hopefully we will be customers one day when im ready to do the move to @PlanetScale

@komran21 About half supabase, and then 1/4 planetscale and 1/4 AWS They essentially get neon branches with our product but get to keep their prod DB on their existing service So for your case you'd get the best of both worlds :)

20 minutes after a demo call w/ a series A co >They want to roll Ardent out across their entire company >They wanna invest Life is good

@jarredsumner damn i just finished learning zig just to work on bun

My team is hiring to make Bun and Claude Code faster. Runtime internals, epoll/kqueue, parsers, JS engine work, long-running agent loop perf. Years of production C/C++/Rust required. $320k - $485k salary. SF in-person. job-boards.greenhouse.io/anthropic/jobs…

I think that eventually every company in the world will build their own version of Shopify River, Ramp Inspect, Stripe Minions. We're building @typedotcom so that you don't have to build it yourself, you can just sign up for type and use it there off the bat

The beauty of @typedotcom is that we are not trying to make our own harnesses. You can use codex, openclaw, and claude code all side by side with your colleagues. The existing skills and agent stacks you're using should work with type just fine. We plan on eventually adding @cursor_ai, @NousResearch 's Hermes and many other harnesses. We want type to be the place where you can experiment with different harnesses and stacks. Want to try @garrytan's gstack on codex before committing to it locally? spin up an agent on type and try it out. By having your AI teammates work in the open people can learn from others, share skills, build automations, etc. It also tends to cut down on people working on the same things. You can spin up a thread to work on something and then move it locally. and @mention your colleagues that you're working on it. It feels like the obvious way to work in the future I like to pitch @typedotcom as the ide for non-technical teams, although I have been using it to build features more and more. (@conductor_build and @superset_sh are still better for feature development and likely always willl be). We still have a ton to build !

@RhysSullivan just the backend*

@RhysSullivan for one, i don't know which version to use. effect-smol? v4? v3? Should I use it on the backend and the frontend? just the frontend?

With agents writing code I’ve become more bullish than ever on Effect and the problems it solves If you’re not using it, I’m curious on why and what you’re using instead to solve the same challenges - interested to see what the ecosystem looks like

@zeeg @bleuonbase @zeeg are you still only using Claude Code?

@bleuonbase Yeah I only use opus at this point

turns out all my opinions were from using codex

@GrowlerEnjooyer @jarredsumner compiler errors are no joking matter, @GrowlerEnjooyer

@komran21 @jarredsumner You're joking, right?

4000 commits and 960,000 LOC. 3 compiler errors until maybe it compiles & panics immediately

@niieani @lassevjl So i can use all eslint rules on oxlint?

@komran21 @lassevjl Biome made the mistake of diverging too far from norms established by eslint. If you know eslint, you know oxlint. Biome has a significant learning curve.

Oxlint >>>>>>> Biome

I really want to use @getsome_air but for some reason worktrees don't work. I selected "New task in git worktree" in the beginning and it just did everything from my root workspace, so now i have a bunch of unrelated changes in my root workspace. It didn't do any setup command or run command configuration. Now I'm getting this git too old error. I've been using git worktrees for like months and now I have to update git? why? plzzz air plz work or just give me a way to manage multiple worktree conversations in webstorm while also having a sane understanding of the code it's writing

@forgebitz This is exactly what we’ve built at @typedotcom I can give you an invite code

what is a good company-wide ai chat app i don't want to lock into one llm and have to switch every 3 weeks just give everyone access to the best models + some shared skills/markdown files

@KhalidWarsa How do you setup CC to do this? Or do you literally copy and paste in between?

I think I’m gonna downgrade my $200/mo Claude max plan and use it for planning, architecting, and reviewing. Then have DeepSeek do the actual implementation.

@VictorTaelin 🤔 what if each harness is unique to the code base?

seriously, working with AI is MISERABLE for one and only one reason: having to re-explain the same thing "oh yeah this new session obviously doesn't know what proper case trees are, so let me explain it for the 5000th time in my life" I'm tired AGENTS.md doesn't solve this because it is impossible to fit the entire domain knowledge without nuking the context - it would be 1m+ tokens worth RAGs don't solve this, the agent won't search unknown unknowns SKILLs don't solve this unless I keep like a collection of 1750 skills with specific cuts of domain knowledge for each possible subset of my domain that I might need in a given chat, but that's a lot of manual work recursive LLMs or whatever don't solve this for the same reason, you can't dump a domain book and expect the AGENT will magically guess that it is supposed to search for a specific bit knowledge. unknown unknowns fine tuning doesn't solve this (OSS models suck and OpenAI / Anthropic gave up on user fine tuning) I honestly think a good product around fine tuning on your domain would be a major hit and an underdog lab should take this opportunity

@kskrygan rooting for intellij

After the Cursor + xAI deal, there are almost no major IDE vendors left that aren’t backed by Big Tech cloud. Welcome to the new money funnel for devs: Antigravity → Google Kiro → AWS Copilot + VS Code → Microsoft Cursor → xAI Almost.

@kskrygan ideally i could bring my codex harness to it

@kskrygan I want a conductor/superset like experience but being able to understand the actual code changes. I dont care about price

Would you be interested if JetBrains releases a totally local AI agent, working 100% on your laptop, using our code insight engine and deeply integrated into the IDE? Yes, it will be probably 1 month behind the very recent frontier models, but no token blood bath anymore WDYT?