AndreaMic

Design Technologist @ frog

[AndreaMic @AndreaMfrog]

https://t.co/ZMUCDzblkR

[AndreaMic @AndreaMfrog]

@orca_build @renanbs82 Thanks @orca_build and @renanbs82, can’t wait to test it!

[AndreaMic @AndreaMfrog]

@renanbs82 @orca_build On mobile aswell?

[AndreaMic @AndreaMfrog]

@orca_build Chat UI for other major harnesses, such as opencode

[AndreaMic @AndreaMfrog]

What about using Jev to attach the only tool schema necessary for the task based on prompt classification?
might this solve the MCP overhead problem?

[AndreaMic @AndreaMfrog]

@MattPRD @Muse Would love to test it

[elie @eliebakouch]

i'm still in shock with this post tbh, it's the same swarm behavior (and model?) that led to the hugging face hack. including it in the investigation is just common sense, leaving it out is dishonest

blaming the lack of a framework for not disclosing this is absurd

i'm very surprised not more people are shocked by this

[AndreaMic @AndreaMfrog]

@samsheffer I found Spark to provide more thorough and accurate results than ChatGPT Work lately, especially after 3.7 drop. However, the experience with the app is not at the level. Crashes often, messages do not get sent, lags when having a large prompt. Can give further details if needed

[AndreaMic @AndreaMfrog]

@ivanfioravanti Ciao Ivan, sono in Cina attualmente e riesco ad accedervi (anche senza vpn). Se hai altre domande fai pure.

[AndreaMic @AndreaMfrog]

@thsottiaux Sincerely, I think price-to-performance is the main thing. When you look at competitors, there is no other alternative that offers this much usage, and these strong models (especially after luna’s discounts) at this prices.
Additionally, I really wanted to try computer use.

[AndreaMic @AndreaMfrog]

@antirez la prossima settimana usciranno i pesi di Qwen 3.8 27B.
Potrebbe essere la prima opportunità per avere coding agents utilizzabili in standard consumer hardware.
Considerando anche i recenti tentativi di rendere DS4 multi-provider, pensi che ci farai qualche pensierino?

[AndreaMic @AndreaMfrog]

Writing my dnd campaign as a knowledge base with OpenAI Codex voice mode while having a swim

[AndreaMic @AndreaMfrog]

I was long a fan of anthropic and their coding models and harnesses, despite the community @antirez @thsottiaux suggestions to check out Codex aswell. I finally did, and there’s no game in most of directions: economics, speed and performance.

[AndreaMic @AndreaMfrog]

@KisekiyaCodes @leouiux The assets 🥰

[AndreaMic @AndreaMfrog]

@leouiux https://t.co/enQ5utLsi2 it’s a bit old tho

[AndreaMic @AndreaMfrog]

Just used gpt work to generate a design poster for a short movie directed by my friend. GPT extracted the styles from the references, elaborate it to match the target content, and sent a custom prompt to image 2.0 that produce a wonderful, unique art.

[AndreaMic @AndreaMfrog]

@leouiux Another banger

[AndreaMic @AndreaMfrog]

@tpitre @nlevin @figma Congrats! You single handedly revolutionised the software this year! You deserve it

[AndreaMic @AndreaMfrog]

@antirez Aura

[antirez @antirez]

Testing Claude Fable. First 30 minutes: very good vibes, but too little time to judge. Now going to make pasta with pesto alla trapanese, re-watch Matrix with my 13yo wonderful girl, and then I'll do more testing. I'll keep you posted.