Onur Solmaz

The models, they just wanna work. They want to build your product, fix your bugs, serve your users. You feed them the right context, give them good tools. You don’t assume what they cannot do without trying, and you don’t prematurely constrain them into deterministic workflows.

@0oAstro @OpenAINewsroom @astral_sh openai runs on python since years...

@OpenAINewsroom @astral_sh bun -> anthropic now, uv -> openai hmm so both have decided what language they are gonna use.

We've reached an agreement to acquire Astral. After we close, OpenAI plans for @astral_sh to join our Codex team, with a continued focus on building great tools and advancing the shared mission of making developers more productive. openai.com/index/openai-t…

Called it x.com/OpenAINewsroom…

Pro tip: tell AI to "explain in plain language" until you understand what you are reading Codex has a tendency to give the full picture, but overcomplicates the response in the process I just use "plain lang" or "plainer lang" as a prompt, it works every time

Thing that codex (and most other models) do that makes me very unhappy { "type": "X", "kind": "Y", ... } And they are so confident too?! Bro we don't use synonyms in our schemas...

This looks extremely cool

I agree with that, I guess I reacted because the statement was too general with skills/specs It reads broader than you intended “Number of additional instructions required to make the model perform better in a harness has been reducing since claude code first came out, and will keep reducing drastically” I guess this is the point but doesn’t sound as good :)

@onusoz I'm saying that most of the "tricks", "hacks", and "skills" and whatnot won't be needed in the future, just like most of the tricks and hacks and feedback loops we used 6 months ago aren't needed anymore.

Lately, whenever I open this app and see the latest tricks, and hacks, and notes, and workflows, and spec here and skill there, I can't help but think: All of this will be washed away by the models. Every Markdown file that's precious to you right now will be gone.

Yes. What is the assumption here? We won't need any write-up in a repo, and AGI can construct that information purely from code? Our monorepo documents all the procedures inside company, like onboarding, testing, releasing, secret management, local development, how to use hubspot, how to debug, how to do code review, and so on...

@onusoz Do you need the same Markdown files when onboarding a new human? Do you need the same skills and MCPs?

@vincent_koc @NousResearch the more the merrier! bring it on :)

@NousResearch 💔 seems we can't be friends anymore 😢

hermes claw migrate

@walpepe1 @zeddotdev then you just need to set nemotron in the harness you use, like opencode changing the model and other configs while on discord is something I still didn’t have time to look at, but once you set a default config, acpx will use that while driving the harness

@onusoz @zeddotdev Nemotron runs via Ollama (OpenAI-compatible API). Any harness that supports OpenAI-compatible endpoints can run it. We're running nemotron-3-super:latest (83B) locally on MacBook M5 I hope this is what you meant by it ollama.com/library/nemotr…

We will support ACP *and* Codex App Server* protocol (CASP) so you get native Codex-like support, and you can use all the others with native ACP or @zeddotdev’s compatibility shims If Anthropic develops their own protocol, we will support that too! The more interoperability and options, the merrier!

Entire world > One company Even in the age of AI

@walpepe1 @zeddotdev what harnesses run nemotron? can you share link?

@onusoz @zeddotdev will we get acp for nemotron

@cnakazawa @badlogicgames wait until he is wearing the lobster suit 💀

I think we are in a bubble.

@bensen @ThePrimeagen 💀

@onusoz @ThePrimeagen so, what about AUDHD?

.@ThePrimeagen made a video about token anxiety, and not being able to focus on one thing My mental model for this is, AI agents cause a shift in the "autism/ADHD spectrum" if you have ADHD, with agents you get Super ADHD if you have autism, with agents you end up mid spectrum or with ADHD this is not scientific of course, just a cultural observation based on what the current memes for these conditions are beside the impact on focus, there is also the economic/competitive pressure, following the realization that anyone could implement the same ideas you are having, so you must be quick this is basically "involution", or 内卷 (Neijuan) in chinese checks out because 996 started to become a meme in SF some time in the last year self-restraint, attention budgeting, and high-level decision making have never been more important if you are in your 20s and have problems with this, I recommend picking up Zazen meditation and yoga every morning, spend 30-40 uninterrupted minutes not doing anything with upright posture, no sounds, just let your brain simmer it helped me in my 20s, I'm sure it will help you too

@mr_r0b0t @BradGroux @BunsDev I deduce this from his recent post, but it might be that your individual profile is being flagged try making a non-anon account free from crypto content and communicate over that

@BradGroux @BunsDev @onusoz There was no mention of crypto 🦧

Request for comments skillflag: A complementary way to bundle agent skills right into your CLIs tl;dr define a --skill flag convention. It is basically like --help or manpages but for agents acpx already has this for example. you can run npx acpx --skill install to install the skill to your agent It's agnostic of anything except the command line It only defines the CLI interface and does not enforce anything else. If you install the executable to your system, you get a way to list and install skills as well Repo currently contains a TypeScript implementation, but if it proves useful, I would implement other languages as well Specification below, let me know what you think! I still think something is missing there. Send issue/PR github.com/osolmaz/skillf…

Agent etiquette is already a thing. This is trending on HN now Don't share huge raw LLM output unedited to your colleagues, it's rude. Your colleagues are not LLMs Either ask the agent to "summarize it to 1-2 plain language sentences", or paraphrase yourself Whenever it is not coming from your brain and instead from AI, always quote it with > to make it clear - even when it is short Respect your fellow humans' attention PSA at stopsloppypasta dot ai

@andyrewlee @ai4k12 thanks for sharing!

@onusoz there's already @ai4k12 and CSTA K-12 CS standards! combining existing standards with ample opportunities to build intuition around using AI models is a good start.

Agent/AI literacy will be a primary school subject in the next 3-5 years How to use and work with agents is going to supersede most other subjects in importance Similarly, robot literacy will follow in 5-15 years