polarzero

software engineer, contributing @amontlabs, prev. @primodiumgame.

[polarzero @0xpolarzero]

Just released evmstate, a TypeScript library to trace & visualize state changes on EVM chains, with human-readable labels (variable names, mapping keys, decoded values, etc).

[polarzero @0xpolarzero]

You can follow the slop fest from this cozy home page: https://t.co/8xO54imPtC

[polarzero @0xpolarzero]

@lennox_saint @CodeTiwg https://t.co/MI6SZ2BKp4

[polarzero @0xpolarzero]

@grok does nanocode reuse the official claude code so model is orchestrated by the harness? Could it be integrated into T3 Code? Do you see any fit inside https://t.co/gjGhL4OhFF, especially as transport for sharing conversations alongside a forked vm, multiplayer coding, or anything else relevant that it specifically enables?

[polarzero @0xpolarzero]

Free and open source, for macOS
https://t.co/MI6SZ2Bczw

[polarzero @0xpolarzero]

I switched to T3 Code with Claude as my daily driver, so I ported lastdrop to it.

lastdrop gives Claude omp-like subagents ("reviewer", "advisor"…) on every other plan I pay for, as if they shipped with the harness:
- Claude calls them through its own Agent tool, the format it's trained on;
- same agents in the chat and in background tasks;
- each task takes the account with the most usage left.

Works with Claude app as well.

[polarzero @0xpolarzero]

@dabit3 @dabit3 is this local or cloud? I couldn't find a way around Apple's terms of maximum 2 macos vms running in parallel, and cloud you can only rent a mac for minimum 24h

[polarzero @0xpolarzero]

@vista8 Hey, just found out about this post, thanks for sharing lcu!

I made another tool for developers using Claude Code you might be interested in.

https://t.co/MI6SZ2BKp4

[polarzero @0xpolarzero]

@dotey Hey, just found out about this post, thanks for sharing lcu!

I made another tool for developers using Claude Code you might be interested in.

https://t.co/MI6SZ2Bczw

[polarzero @0xpolarzero]

Obviously agent-desktop looks really cool

[polarzero @0xpolarzero]

You can do this with silo as well: https://t.co/gjGhL4OhFF.

You get both macos and linux computers for your agents, with builtin codex computer use, which until proven otherwise is the best computer use out there.

Still heavy wip and I’m working on a better app that silo will fold into tho.

@mdlahfir any benchmarks on how agent-desktop compares with codex cu?

[Lahfir ▲ @mdlahfir]

if you're on a mac and want computer use without the agent touching your personal stuff, here's how i do it

run agent-desktop inside a macOS VM. Apple's Virtualization framework boots a full guest in a window, and the agent only sees what's inside it.

you keep working on your mac and your agents runs inside the VM

[polarzero @0xpolarzero]

Use any model as a built-in subagent in Claude Code, with a subscription or API key.

Claude calls them like its native subagents, and lastdrop spreads usage across your accounts.

https://t.co/MI6SZ2Bczw

[z80 🍋 @0xz80]

I have decided to take the leap out of crypto and find a new job in AI

If you, or anyone you know, is hiring, I would love to chat!

let’s cook 🧑‍🍳

ps. I’m also open to crypto-related roles if they’re interesting

[polarzero @0xpolarzero]

Claude code is the best harness by far.

2 things I was missing I now have as mods built on top of the app:
- (a good) computer use: https://t.co/cpeMCctQR8
- subagents (from multiple claude subs and other providers): https://t.co/0AgGQApvmF

(slop motion showcases attached)

[polarzero @0xpolarzero]

@GuillaumeLample Bravo pour la release! Des plans pour open-source le modèle dans le futur?

[polarzero @0xpolarzero]

@MistralAI Closed source?

[polarzero @0xpolarzero]

@injaneity @francedot @pkyanam @NOOROU Thanks for answering. I'm not sure I recorded these benchmarks, these were vibed on a few samples and few iterations, just enough to get an idea. I was more expecting to see yours ahah

[polarzero @0xpolarzero]

@francedot @pkyanam @NOOROU @injaneity Last time I benchmarked cua was less accurate than codex computer use, is it not the case anymore? I even made a library for using codex cu inside other harnesses for this reason

[polarzero @0xpolarzero]

We crossed 700 stars.

Keep the issues coming, and feel free to open a PR if you believe you have a useful feature or fix to contribute!

[polarzero @0xpolarzero]

I extracted computer use from the ChatGPT app to plug it into any harness, because it's great.

Completely decoupled from ChatGPT auth and built-in browser. For macOS and Linux.

https://t.co/HLU60T4ISd

[Guillaume Lample @ NeurIPS 2024 @GuillaumeLample]

Today, we are launching a preview of our new model, Mistral Large 4 (ML4), aka le Chonk 🐈.
ML4 is a 1T-parameter model with 49B active parameters, trained natively with multimodal capabilities.

It is at the frontier of open models, and by far the strongest open-weight model from the US or Europe.
The RL run behind this preview is still in flight and shows no sign of saturation -- we will release a final version before the end of the month along with the weights of the model.
🧵
1/n

[polarzero @0xpolarzero]

https://t.co/cpeMCctj1A

[polarzero @0xpolarzero]

Btw with lcu you can get claude to control... claude.

Computer-use agents usually don't allow a harness to control itself but we do.

[polarzero @0xpolarzero]

LCU now works properly with Claude Code (both app and CLI).

The first time Claude touches an app, you get an approval panel right in the session.

You can also review and edit your always-allowed apps with Touch ID, no Codex app needed.

https://t.co/cpeMCctj1A