Pranav

Building Audrey AI ( | ex @intercom @JuneDotSo @getsolidroad

[Pranav @pranav7]

domux — https://t.co/loxB8ZN6uP

[Pranav @pranav7]

@shadcn Nothing really! Terminal still wins for me

[Pranav @pranav7]

todolist is a simple tracker to remember what you were doing on that workspace.

That's it, feedback welcome!

[Pranav @pranav7]

domux — https://t.co/loxB8ZN6uP

[Pranav @pranav7]

Two main parts to domux, the switcher and the todolist.

switcher is a tmux session switcher, but rich with details like the branch a worktree is on, a name for your session, preview of the tasks attached to that session, shows the status of what AI agent attached to that session

[Boris Cherny @bcherny]

Dogfooding Opus 4.7 the last few weeks, I've been feeling incredibly productive. Sharing a few tips to get more out of 4.7 🧵

[Wes Bos @wesbos]

Claude Code leaked their source map, effectively giving you a look into the codebase.

I immediately went for the one thing that mattered: spinner verbs

There are 187

[Lydia Hallie ✨ @lydiahallie]

Claude Code now supports agent teams (in research preview)

Instead of a single agent working through a task sequentially, a lead agent can delegate to multiple teammates that work in parallel to research, debug, and build while coordinating with each other.

Try it out today by enabling agent teams in your settings.json!

[DHH @dhh]

We're all still fumbling in the dark with agents, but some patterns are emerging (and some disappear quickly again!), and it's our role as software makers to make use of it all along the way. Here are some quick notes from one of those internal sessions sorting it all at 37s.

[Pranav @pranav7]

Finally got a setup I'm happy with for running multiple Claude agents locally. Long-running git worktrees + tmux + fzf session switcher + Claude hooks for status updates. Terminal-native parallel AI dev.

[chirag @mrnacknack]

https://t.co/M2BtVDdhkM

[Pranav @pranav7]

Tyops are dead.

[Pranav @pranav7]

the dans know what they are doing
https://t.co/jp3hxQF2Wu

[Pranav @pranav7]

keeps getting better

[Pranav @pranav7]

One of the good ones I've seen so far

[Pranav @pranav7]

whatever this means

[Pranav @pranav7]

superwhisper is the kind of software that naturally just stuck in my workflow. I get lazy when I have to type, but not when I'm speaking

has massively helped in sharing a lot more context when prompting an ai agent. i would just speak to it like i'm speaking to another engineer

[Pranav @pranav7]

My focus remains heavy on new ground, which we have a lot at @getsolidroad given we're so young. Then let AI agents take the patterns forward.

One hand in the mud, the other delegating at scale.

[Pranav @pranav7]

A friend asked me recently (rather sarcastically, but a legit question still):

> “If agents are doing all the work… what do YOU do?”

[Pranav @pranav7]

If it’s an established pattern and the codebase has a decent example to point to, agents are very good at recognizing and reusing it. At that point, I can point an agent to the example and say: “Here’s how we did X, now let's build Y.” And it’ll usually land close to perfect.