Entries for February 2026

  1. Portrait of Onur Solmaz
    @onusoz · /2026/02/28· View on
    I created a semi-automated setup for ingesting X posts into my blog, and it works pretty well! I own my posts on X now Posts are scraped while I browse X using @kubmi's xTap and get automatically synced to my blog repo. Posts saved as jsonl are then converted to jekyll post pages according to my liking I reproduced the full X UI/UX, minus stuff like like count. Now all my posts are backed up in my blog, and they are safe even if something happens to my account here! The posts are even served over RSS! So you can subscribe to it without going through X! Reply if you want to set this up for yourself, then I will put some effort into standardizing it
  2. Portrait of Onur Solmaz
    @onusoz · /2026/02/28· View on
    Agentic Engineering is a newly emerging field, and we are the first practitioners of it. Currently there is a lot of experimentation going on, and there is a large aspect to it that is more ART then engineering For example, @steipete says "you need to talk to the model" to get a feel. a lot of work around refining how an agent feels like, sounds like psychology. this part is crucial and should not be ignored, looking at openclaw's success but then there is the hardcore engineering part of it, e.g. Cursor creating a browser or anthropic a C compiler from scratch fully autonomously and there is a whole other dimension of how to teach all software developers this new discipline, lest they be jobless what is obvious is that everybody is trying to grasp for things in the dark and that we need more RIGOR. the art/psychology aspect of it aside, we need solid engineering fundamentals the "thermodynamics" of this new discipline will most likely be formal verification and program synthesis. we might have some breakthroughs that will make certain things clear. the products of it will most likely include a new programming language optimized for agents and the speed of inference moreover, it would be foolish to thing agentic engineering is limited to software. it will penetrate every aspect of the economy, bits AND atoms. it will over time evolve into the engineering of managing robots @simonw is now leading in collecting very useful info from the practitioner's point of view, I highly recommend you to follow this thread let's formalize our new field together!
  3. Portrait of Onur Solmaz
    @onusoz · /2026/02/27· View on
    this is an insane deal @greptile, and probably an unsustainable one depending on your team, getting a similar service in codex github review credits is in my head 3~5x more expensive go get a greptile sub everyone while the free lunch lasts
  6. Portrait of Onur Solmaz
    @onusoz · /2026/02/27· View on
    mfw codex tries to create a backward compatibility layer to a schema that it created 2 turns ago before compacting there is no v2 bro what are you doing...
  7. Portrait of Onur Solmaz
    @onusoz · /2026/02/27· View on
    Note that this is currently in beta, but will ship in a couple of hours
  8. Portrait of Onur Solmaz
    @onusoz · /2026/02/27· View on
    Claude Code / Codex in Discord threads is shipped now! To enable, copy and paste this to your agent: ``` Enable feature flags: acp.enabled=true acp.dispatch.enabled=true channels.discord.threadBindings.spawnAcpSessions=true Then restart. After restarting: Start a codex (or claude code) discord thread using ACP, persistent session, just tell it to write a haiku on lobsters to initialize acpx for the first time ``` You may need to nudge your agent to “continue” after restarting The first implementation is very barebones, I have made it work in a clean way and merged. In a codebase like openclaw’s, it’s better to develop incrementally Please send any issues my way. I am already aware of some and working on to fix them
    @steipete·
    And another super cool feature: codex/claude code can now be first-class subagents via acp! https://t.co/xq0SprBi5A
  9. Portrait of Onur Solmaz
    @onusoz · /2026/02/26· View on
    an agent is an LLM in a loop with tool call a claw is an agent in a messaging app
  10. Portrait of Onur Solmaz
    @onusoz · /2026/02/26· View on
    Update acpx to the latest version 0.1.13 npm i -g acpx@latest There was a bug that caused an unnecessary hang on calls to acpx <harness> prompt, should be fixed now
  12. Portrait of Onur Solmaz
    @onusoz · /2026/02/25· View on
    MIT License on everything from now on. It doesn't make sense to use anything else, except for a few large projects that hyperscalers exploit and not give back If you were making money from a niche app, open source it under MIT License If you had an open source project with GPT, convert it into MIT Extreme involution is about to hit open source. Code is virtually free now. If you want your projects and their brand to survive, the only rational strategy is to remove all barriers in front of their adoption, and look for other ways to survive
  13. Portrait of Onur Solmaz
    @onusoz · /2026/02/25· View on
    I spoke in absolute terms, I meant to say *feels*
  14. Portrait of Onur Solmaz
    @onusoz · /2026/02/25· View on
    This. Agent Experience first. Agent Ergonomics. we need to get used to these terms
    @_andydeng·
    We have long paid close attention to improving user experience (UX) and developer experience (DX) when building apps and tools. There is no doubt that we are now entering an era where the apps and services we produce must also—if not primarily—cater to the needs of a new type of consumer: AI agents. This means that from now on, we need to think about delivering good AX (Agent Experience). We have seen this trend forming ever since the birth of MCP and, later, the popularity of Skills. A recent blog post from the Next.js team discussed the necessity of exposing more information to agents within development tooling, allowing coding agents to make better decisions based on a more complete awareness of errors and outputs. It is living proof that the software we build needs to adapt to this new type of user. With OpenClaw becoming the embodiment of the powerful personal agent almost overnight, we are seeing platforms dedicated to agents, like MoltBook and ClawNews, burst onto the scene. Moreover, the simplicity of OpenClaw (and the underlying pi agent it uses) regarding tool calling has boosted the practice of packaging web services into simple, local CLIs. This form of interface had long fallen "out of fashion" for common users due to its abstract UX, but it is now regaining popularity because of its simplicity and token-efficiency for AI agents. Even though we might not have fully figured out where these agent-oriented web platforms will lead—they might simply be slop art, or they might one day catalyze meaningful agent self-growth and collaboration—there is no doubt that what an agent needs when interacting with the web, apps, and services is vastly different from what human users need. ## The Loop Humans and AI agents both rely on a loop when consuming information and completing tasks. We, the humans, access information mostly via apps and websites. We open an interface, we read, click, type, think, and repeat. AI agents, on the other hand, receive initial prompts from humans and then basically stay in a loop of communicating with LLMs and calling tools to obtain more context and data until a conclusion is reached and they report back. (The human element might become less important or even unnecessary once these agents become highly autonomous.) ![HBk8S_HbAAAh5rE.png](media/2024716661576913092/HBk8S_HbAAAh5rE.png) ## The Beauty of Files and CLIs Tool calling is clearly the most vital mechanism an agent relies on to interact with the external world. As demonstrated by the Pi agent, the tools an agent needs generally boil down to just two categories: file operations (read, write, edit) and command execution (to consume services or obtain data). Humans prefer rich, smooth interactions and visually appealing UIs to efficiently consume information, produce artifacts, and complete tasks. For agents, however, efficiency looks entirely different: the more straightforward the path between request and result, and the fewer tokens the process imposes, the better. This is why the progressive disclosure of data employed by the Skills system is highly favored over MCP. The input-output efficiency of command-line tooling has become the easiest way for agents to gain knowledge of the outside world. ## The Middleware Obstacle For agents, traditional webpages and app interfaces are essentially obstacles. There are already countless attempts to help agents navigate existing webpages or operate mobile apps. Google recently released WebMCP in an attempt to lower the barrier for agents operating on websites. While this looks very promising as a unified approach catering to a massive number of legacy websites, a browser still sits in the middle, forcing agents to interact with interfaces built for human consumption. Understandably, these middleman approaches will likely persist for some time, as there is no unified mechanism to retroactively fit the old web into the new world. Beyond the asymmetrical demands between humans and agents when accessing information, we cannot ignore the fact that a large number of websites actively and aggressively block autonomous agents. This seems logical for social and UGC platforms like X—after all, we aren't quite ready to socialize with agents yet. However, these protective measures vividly demonstrate that the web, its apps, and the mindset behind them were fundamentally not designed for agents. ## APIs and CLIs - The Agentic Way It has become clear that for agents, direct access to functionality—even in the form of raw, low-level APIs—is far superior to wiggling through complex UIs. In that sense, many companies currently selling apps and tools to human users will eventually pivot to selling APIs or CLIs to agents. If the business value exposed through these APIs isn't sophisticated enough to prevent an agent from replicating it with relatively little effort, the business model itself might not survive as LLMs evolve. This is exactly what Karpathy discussed in a recent tweet: > "99% of products/services still don't have an AI-native CLI yet." -- Andrej Karpathy In 2026, new and existing services will become hyper-aware of agent ergonomics and will start offering better experiences via CLIs or streamlined APIs. When building products or producing digital artifacts, developers will always have to consider "the other type of user." For all apps, functionalities currently exposed via human-oriented UIs are destined to be transformed into CLI arguments or API parameters. Either that, or someone will build an AI-native alternative specifically to cater to agent needs. Documentation that teaches users how to navigate specific workflows will also start to be complemented with agent Skill specifications. This transformation will be fast for some categories of apps and slow for others. Applications that employ proprietary formats, rely on highly complex data manipulation logic, or fundamentally require human intuition will probably retain their current forms for a while. Eventually, however, agents will find a way to become the primary users of even the most complex software. Looking further ahead, sensors bridging the physical and digital worlds—or devices providing pure physical utility—can also benefit from this transformation. Even when agents gain a physical form, like a humanoid robot, and possess the skills to navigate environments designed for humans, an agent-friendly digital interface will still be superior in certain contexts. Imagine a humanoid operating a rice cooker: it absolutely needs vision and motors to lift the lid and pour the rice. But when it comes to setting the timer or selecting the cooking mode, a direct API exchange will always be more efficient than using a camera to look for a digital panel and using a robotic finger to press the buttons. ## The Inevitable Standard of AX We are rapidly approaching a tipping point where AX will require universal standards. Just as we have spent the last decade obsessing over SEO and performance metrics to help search engines parse our sites, we will soon need standardized protocols for agent interaction. Whether these emerge as universally adopted CLI wrappers, predictable API architectures, or even decentralized agent-centric standards like ERC-8004, a structured framework is inevitable. The platforms and developers who define these protocols will ultimately dictate the next decade of digital infrastructure. Ultimately, this transition from UX to AX is not about removing humans from the equation; it is about removing friction for the increasingly autonomous tools we employ. When our software is optimized for the entities that process data the fastest—our agents—our own capabilities are amplified. We are moving past the point where an agent-friendly interface is just a clever feature. Very soon, building for agents won't just be an alternative to building for humans; it will be the prerequisite for participating in the digital ecosystem.
  15. Portrait of Onur Solmaz
    @onusoz · /2026/02/25· View on
    OpenAI nerfed GPT 5.3 Codex xhigh. We independently reported the same thing at @TextCortex today I'm looking forward to deploying open models and putting an end to this paranoia
  16. Portrait of Onur Solmaz
    @onusoz · /2026/02/25· View on
    "academics"
    @_vgnsh·
    It's pretty disingenuous to specifically set up an application in all ways it isn't meant to be then claim they red teamed it. - openclaw is a personal assistant. It isn't meant to be thrown in a discord channel with a dozen users and others agents with elevated tool access. - methodology is severely lacking. No comparison with other personal agent systems like Claude Cowork. Repeatedly doing something not meant to be done with openclaw and claiming "agent is compromised" is a deceptive methodology. - this paper opened up openclaw to moltbook (lol). The researchers opened themselves up to far serious vulnerabilities than what the agent did by itself. - it's just sad to see a poorly structured research and paper being advertised as new and necessary research. And riding on the popularity of openclaw.
  17. Portrait of Onur Solmaz
    @onusoz · /2026/02/25· View on
    In the hall of OpenClaw GitHub repository, I brought my PR before Master @steipete He read it once, then laid it aside "You act," he said, "as if code were not cheap." At these words, I was enlightened I bowed
  18. Portrait of Onur Solmaz
    @onusoz · /2026/02/24· View on
    woah chatgpt web app now has steering, and much more different streaming behavior huge upgrade behind the scenes, must have come up in the last few days
  19. Portrait of Onur Solmaz
    @onusoz · /2026/02/24· View on
    the lobster looks good on acpx
  20. Portrait of Onur Solmaz
    @onusoz · /2026/02/23· View on
    imagine if tarantino were 16 years old now and saw seedance 2.0 95% of videos i saw since the launch for absolute tasteless slop. they are going viral because of ragebait but soon, serious imagineers will start entering the game, and they will learn to shape generation output exactly how they want it's the best time to be young and full of imagination
  23. Portrait of Onur Solmaz
    @onusoz · /2026/02/23· View on
    codex in discord achieved
  24. Portrait of Onur Solmaz
    @onusoz · /2026/02/23· View on
    acpx v0.1.7 is out improvements to json mode and other functionality to make it possible to integrate acpx as a backend into other harnesses, like openclaw
  25. Portrait of Onur Solmaz
    @onusoz · /2026/02/23· View on
    POV: you became a plumber after all, just for agents
  26. Portrait of Onur Solmaz
    @onusoz · /2026/02/23· View on
    @grok what do you think should replace it? what happens to belief when the cost of creating software goes to zero?
  27. Portrait of Onur Solmaz
    @onusoz · /2026/02/23· View on
    another thought i'm having these days is that we need a new philosophy of free software (as in freedom), or an update to it the most psychologically imprinting philosophy is stallmanism, and the philosophy of FSF. it is righteous and strict, and i believed it growing up but GPL and money don't go well together. that's why most of the lasting open source projects today use MIT, Apache and the like. it turns out you can still make a good living with open source. i want to make money, so i never use GPL in my projects and to add another deadly blow to stallmanism, code is cheap now, virtually free does this mean stallmanism is dead? if there is an open source project using GPL that i want to use commercially, i can now recreate it from the original idea and intent completely independent of it (ignoring training data), just like how i can recreate a proprietary service stallmanism was already long-irrelevant. but does this mean we must finally declare it dead? code is free now. what does it mean for open source? what replaces stallmanism?
    @onusoz·
    one effect openclaw had on me is that I've bought a gpu home server, set it up with tailscale and now doing a lot of work through ssh and tmux like i did 10-15 years ago im back on linux, considering buying an android phone again it's time to dream big again and unshackle ourselves from proprietary software. it's time to build
  29. Portrait of Onur Solmaz
    @onusoz · /2026/02/22· View on
    one effect openclaw had on me is that I've bought a gpu home server, set it up with tailscale and now doing a lot of work through ssh and tmux like i did 10-15 years ago im back on linux, considering buying an android phone again it's time to dream big again and unshackle ourselves from proprietary software. it's time to build
  31. Portrait of Onur Solmaz
    @onusoz · /2026/02/21· View on
    In the new release OpenClaw, you can talk to subagents in Discord threads Currently a beta feature so ask your agent to set session.threadBindings.enabled=true Next up: - Telegram, slack, imsg threads - Use ACP to talk to Codex, Claude Code and other harnesses on your machine
    Quoted post
    Quoted post was not retrieved.
  33. Portrait of Onur Solmaz
    @onusoz · /2026/02/21· View on
    openclaw might be the highest velocity codebase in the world, and soon, others will follow as well conflict anxiety is real, it's like trying to shoot a moving target every time. I wonder if our existing tooling will ever solve this problem feel like faster models might. but then the rate of conflict creation is also tied to that. might be unsolvable
  37. Portrait of Onur Solmaz
    @onusoz · /2026/02/20· View on
    Imagine not having to upload skills to 3-4 competing skill registries for each of your projects Turns out we already have a skill registry: npm skillflag lets you bundle skills right into your CLI's npm package, so that you can run --skill install github -> osolmaz/skillflag
  38. Portrait of Onur Solmaz
    @onusoz · /2026/02/20· View on
    Scoop, our open source home news intelligence platform can now translate foreign language into english for free, using on-device models github -> janitrai/scoop
  39. Portrait of Onur Solmaz
    @onusoz · /2026/02/20· View on
    A picture is worth a thousand words, so acpx now has this cute banner Also, updated skillflag tooling so that you (or better, your agent) can just call: npx acpx@latest --skill install acpx
  41. Portrait of Onur Solmaz
    @onusoz · /2026/02/20· View on
    @kepano I would grow my own vegetables if I had equally cheap access to and ownership of land, alas I am disenfranchised Prompting an agent is much easier compared to plowing a fields Farming analogies break when it comes to software https://t.co/CkldO8eWKc
    @onusoz·
    People like the farmer analogy for AI Like before tractors and industrial revolution 80% of the population had to farm. Once they came all those jobs disappeared So analogy makes perfect sense. Instead of 30 people tending a field, you just need 1. Instead of 30 software developers, you just need one Except that people forget one crucial thing about land: it's a limited resource Unlike land, digital space is vast and infinite. Software can expand and multiply in it in arbitrarily complex ways If you wanted the farming analogy to keep up with this, you would have to imagine us creating contintent-sized hydroponic terraces up until the stratosphere, and beyond...
  42. Portrait of Onur Solmaz
    @onusoz · /2026/02/20· View on
    acpx v0.1.5 is out now it is much more feature complete in terms of ACP. your agent can send, queue and cancel messages to Claude Code, Codex, Pi, or ant other coding agent npm install -g acpx@latest
  44. Portrait of Onur Solmaz
    @onusoz · /2026/02/19· View on
    on another note, I do believe AI will play a huge part in families growing up in late 90s, my dad taught me the importance of reading newspapers and being informed of the world. my nickname in middle school was "newspaper boy" for a long time because I read the newspaper in class on September 12, 2001. i was 10 years old then I witnessed the enshittification of media and journalism in the following decades. today, serious journalists are setting up their own boutique agencies and bypassing mainstream media. important news land on individual accounts before mainstream agencies but there is simply too much to consume. something must filter out the noise and digest the info according to the family's preferences i think AI will play a big role in family intelligence. proprietary family heirloom AI, weights fully owned by the family it will be the parents' job to filter out the signal from the noise, and train the AI on what is right and what is wrong for the family. family and friend circles will let their AIs talk to each other and share important information consuming mass media and mass AI will not be enough to survive and prosper in the new world. families will need to be proactive about how they and their children use AI
    @onusoz·
    on ai psychosis 80% of people need to use ai agents in a very sterile and boring way in order not to go crazy majority of the population does not have the skepticism muscle. they don't have theory of mind, and will subconsciously and emotionally associate with machines, while on the surface lying to themselves that they don't especially those that grew up in the us under hardcore consumerism and adjacent cultures you thought 4o addicts were bad? wait a few years, it will get much worse. we will have to regulate all this if you don't want to become a victim of this, make your openclaw SOUL. md as bland as possible. mine knows it's just a tool and this is a subjective view of course. @steipete might disagree with me. his instance feels much more interesting and fun. i truly like that one better but that is exactly the problem for me. i know myself, and i know it is a slippery slope for me. so i self regulate and set up my system accordingly. thankfully, im an adult and my brain has set enough such that any damage would be limited but there is a risk for emotionally vulnerable people, or children, specifically a risk of dissociating and losing touch with reality why do i write all this? because being in this project, i feel responsible, and feel like we should prepare for what is to come
  45. Portrait of Onur Solmaz
    @onusoz · /2026/02/19· View on
    on ai psychosis 80% of people need to use ai agents in a very sterile and boring way in order not to go crazy majority of the population does not have the skepticism muscle. they don't have theory of mind, and will subconsciously and emotionally associate with machines, while on the surface lying to themselves that they don't especially those that grew up in the us under hardcore consumerism and adjacent cultures you thought 4o addicts were bad? wait a few years, it will get much worse. we will have to regulate all this if you don't want to become a victim of this, make your openclaw SOUL. md as bland as possible. mine knows it's just a tool and this is a subjective view of course. @steipete might disagree with me. his instance feels much more interesting and fun. i truly like that one better but that is exactly the problem for me. i know myself, and i know it is a slippery slope for me. so i self regulate and set up my system accordingly. thankfully, im an adult and my brain has set enough such that any damage would be limited but there is a risk for emotionally vulnerable people, or children, specifically a risk of dissociating and losing touch with reality why do i write all this? because being in this project, i feel responsible, and feel like we should prepare for what is to come
  47. Portrait of Onur Solmaz
    @onusoz · /2026/02/18· View on
    I have improved acpx sane defaults When your agent runs acpx codex in a different project, it starts a new session If it tries to run it in a subfolder in your project, it still finds the session in your repo root Also, starting a session needs an explicit `sessions new`, so that it doesn't accidentally litter your project with sessions Tell your agent: Run this and install acpx per instructions: npx acpx@latest --skill show acpx
  48. Portrait of Onur Solmaz
    @onusoz · /2026/02/18· View on
    Your markdown files are executables now Relatedly, your install instructions can be as well. Copy and paste markdown to your @openclaw to install acpx
  49. Portrait of Onur Solmaz
    @onusoz · /2026/02/18· View on
    So who is building actually good open source self hostable discord that supports token streaming now? And who is building an open source version of codex desktop app?
  50. Portrait of Onur Solmaz
    @onusoz · /2026/02/18· View on
    and of course, I've used `acpx codex` to build acpx itself... magical feeling when the tool builds itself
  52. Portrait of Onur Solmaz
    @onusoz · /2026/02/18· View on
    I am a fan of @zeddotdev by this point, it’s currently my daily driver It’s not perfect, but I feel it’s travelling on the right direction at a faster rate compared to other editors
  53. Portrait of Onur Solmaz
    @onusoz · /2026/02/18· View on
    ACP appreciation post Agent Client Protocol by @zeddotdev is extremely underrated right now. We have bazillion different harnesses now, and only one company is working competently to standardize their interface 💪
    @onusoz·
    You know how it's a pain to work with codex or claude code through @openclaw? Because it has to run it in the terminal and read the characters for a continuous session? I have created a CLI for ACP so that your agent can use codex, claude code, opencode etc. much more directly Your agent can now queue messages to codex like how you do it Shoutout to @zeddotdev team for developing the amazing Agent Client Protocol, ACP! I just glued together the pieces Repo: janitrai/acpx npm i -g acpx
  54. Portrait of Onur Solmaz
    @onusoz · /2026/02/18· View on
    You know how it's a pain to work with codex or claude code through @openclaw? Because it has to run it in the terminal and read the characters for a continuous session? I have created a CLI for ACP so that your agent can use codex, claude code, opencode etc. much more directly Your agent can now queue messages to codex like how you do it Shoutout to @zeddotdev team for developing the amazing Agent Client Protocol, ACP! I just glued together the pieces Repo: janitrai/acpx npm i -g acpx
  57. Portrait of Onur Solmaz
    @onusoz · /2026/02/16· View on
    Link to the post: https://t.co/C3Ac0jLFwh
  58. Portrait of Onur Solmaz
    @onusoz · /2026/02/16· View on
    I wrote a deeper blog post about how I built a coding agent 2 months before ChatGPT launched, on my blog "When I made icortex, - we were still 8 months away (May 2023) from the introduction of “tool calling” in the API, or as it was originally called, “function calling”. - we were 2 years away (Sep 2024) from the introduction of OpenAI’s o1, the first reasoning model. both of which were required to make current coding agents possible." Still bends my mind... Link to the post below
  59. Portrait of Onur Solmaz
    @onusoz · /2026/02/16· View on
    Who here remembers the OG Codex launch from 2021 😏 Also, Greg and Ilya in the same room 😭
  62. Portrait of Onur Solmaz
    @onusoz · /2026/02/15· View on
    I'm building a news intelligence platform to be used by my openclaw instance @dutifulbob, SCOOP local first, using local embedding model (qwen 8b) ran into the issue because bob was giving me a repeat of the same news every day. it needed a system in the background to deduplicate different news items into single stories interface is simple, call `scoop ingest...` with the json for the news item. it gets automatically analyzed and added to the pg database running pgvector currently, it's just doing simple deduplication and gives me a nice UI where I can view the story and basically use it as an RSS reader next up: implement custom logic for my preference of ranking. for example, get upvote counts from hacker news and reflect it to the item's ranking on the feed I want this to be fully hackable and adjusted to your preference. It should scale to thousands of news items ingested daily on your local machine, and be able to show you the most important ones Usable by both you and your agent github -> janitrai/scoop
  63. Portrait of Onur Solmaz
    @onusoz · /2026/02/15· View on
    Training all these models of different sizes, on changing datasets and running experiments have also revealed some challenges that I feel profs would never teach at a uni ML program Like how to cleanly keep track of the gazillion runs Yeah I can name them after layer dims and other stuff, but that's to me like trying to remember UUIDs So I ended up choosing iso datestamp + petname, like 2026-02-15-flying-narwhal If anyone has a convention that is easier on the brain and the eyes, I am all ears
  64. Portrait of Onur Solmaz
    @onusoz · /2026/02/15· View on
    I have a GPU now, so I can do ML experiments on @janitr_ai crypto/scam detection dataset - I trained a tiny student BERT (transformer for the nonfamiliar), 3.6 MB ONNX model, still lightweight for a browser extension - Still fully local on your device (no cloud inference) - On frozen unseen holdout data (n=1,069), exact prediction accuracy improved from 77% -> 82% - Scam detection improved: precision 91% -> 94%, recall 55% -> 61% - Scam false alarm rate improved from 1.58% -> 1.21% And models are on huggingface org now, handle is janitr
  66. Portrait of Onur Solmaz
    @onusoz · /2026/02/15· View on
    waiting compilation and execution will soon be the bottleneck again. and we’ll write the entire stack from scratch in a matter of years, because we can Andy and Bill’s law will change and we’ll see incredible performance gains with the same hardware we already have like what @astral_sh is doing to python, but with everything that is slow and has accumulated cruft
    Quoted post
    Quoted post was not retrieved.
  67. Portrait of Onur Solmaz
    @onusoz · /2026/02/15· View on
    we need a protocol for agent <> app interaction something that natively accounts for the abuse factor and let’s agents consume by paying. NOT crypto, NOT visa, something that’s agnostic of the accounting and payment system and then all UIs will be purely for human clicking/tapping + instaban on the first proof of programmatic exploit people will still make agents mimic humans, and every platform will have to invest in more sophisticated bot detection this arms race will just proliferate, but we can at least start by creating legal channels for agents to consume data
    Quoted post
    Quoted post was not retrieved.
  68. Portrait of Onur Solmaz
    @onusoz · /2026/02/15· View on
    I am now training smol bert models on my gpu for @janitr_ai scam detection it's funny how I have to discover everything from scratch. like the models don't even know how to lay out performance metrics in a nice way in the terminal for a human to view and decide during experiments it would by default bombard me with numbers that do not make visual sense. I then created a skill with common sense: - metrics always on y-axis, candidates on x-axis - write without zero and 2 sigfigs,.12 instead of 0.12345 - align the dots - use asterisks to show which alternative is the best: 0-1% difference -> considered equal 1-5% -> * 5-10% -> ** 10-50% -> *** > 50% -> **** visualization skill is in @janitr_ai repo for anyone who is interested
  69. Portrait of Onur Solmaz
    @onusoz · /2026/02/14· View on
    no other occupation has been catapulted from one end of the spectrum (autism) to the other (adhd) in such a short time
  70. Portrait of Onur Solmaz
    @onusoz · /2026/02/14· View on
    I've helped our sales team to build CLIs for some SaaS that we pay for on their side We are letting our agents call the APIs sensibly and not abuse things Calling a backend is a verifiable task. It takes a single prompt to codex to create a CLI for any API We are early, but everybody will start doing this very soon. Incumbent SaaS will face a choice. Either: (1) embrace agents and the new medium of consumption and change their business model into a pay-per-use API like X is doing, or (2) keep it purely for humans Those that choose (2) will get wiped out of business. And I fear many will choose (2) Which means you can just copy an incumbent's product, make it consumable through a CLI, and make a lot of $$$
  72. Portrait of Onur Solmaz
    @onusoz · /2026/02/13· View on
    The good thing about @levelsio and others flagging AI replies in public is that they are perfect annotations for the open @janitr_ai dataset Just searching “blocked for ai reply” yields hundreds of samples for seed data
  73. Portrait of Onur Solmaz
    @onusoz · /2026/02/13· View on
    github added a new agents tab between pull requests and actions. single glance and i don't feel like giving it a try at all
  74. Portrait of Onur Solmaz
    @onusoz · /2026/02/13· View on
    *puts on schmidhuber hat* well ackshuaally i created the first coding agent back in 2022, 2 months before chatgpt launched jokes aside, it's super cool how I have come full circle. back in those days, we didn't have tool calling, reasoning, not even gpt 3.5 it was codex THE CODE COMPLETION MODEL and frikkin TEXT-DAVINCI-003 for some reason, I did not even dare to give codex bash access, lest it delete my home folder. so it was generating and executing python code in a custom jupyter kernel you can even see the approval gate before executing. I was so cautious, for some reason, presumably because smol-brained model generated the wrong thing 80% of the time. definition of being too early Antique repo:
  75. Portrait of Onur Solmaz
    @onusoz · /2026/02/13· View on
    you can order bubble tea in qwen in china? @TextCortex when berlin döner in zenochat? https://t.co/O4I950ltEO
  76. Portrait of Onur Solmaz
    @onusoz · /2026/02/13· View on
    it happens these days that I am telling an model to prompt another model. the reason is often the model I am using (opus) is a bad designer. not only it's not a bad designer, it is a bad reasoner and it doesn't understand from the context why it's made to ask another model so I have to create a skill to prevent it from biasing the smarter model (codex) with its bad suggestions
  77. Portrait of Onur Solmaz
    Onur Solmaz · Log · /2026/02/13

    I built a coding agent two months before ChatGPT existed

    I built a coding agent back in 2022, 2 months before ChatGPT launched:

    It’s super cool how I have come full circle. back in those days, we didn’t have tool calling, reasoning, not even GPT 3.5!

    It used code-davinci-002 in a custom Jupyter kernel, a.k.a. the OG codex code completion model. The kids these days probably have not seen the original Codex launch video with Ilya, Greg and Wojciech. If you have time, sit down to watch and realize how far we’ve come since August 2021, airing of that demo 4.5 years ago.

    For some reason, I did not even dare to give codex bash access, lest it delete my home folder. So it was generating and executing Python code in a custom Jupyter kernel.

    This meant that the conversations were using Jupyter nbformat, which is an array of cell input/output pairs:

    {
  "cells": [
    {
      "cell_type": "code",
      "source": "<Input 1>",
      "outputs": [
         ... <Outputs 1>
      ]
    },
    {
      "cell_type": "code",
      "source": "<Input 2>",
      "outputs": [
         ... <Outputs 2>
      ]
    }
  ]
}

    In fact, this product grew into TextCortex’s current chat harness over time. After seeing ChatGPT launch, I repurposed icortex in a week into Flask to use text-davinci-003 and we had ZenoChat, our own ChatGPT clone, before Chat Completions was in the API (it took them some months). It did not even have streaming, since Flask does not support ASGI.

    As it turns out, nbformat is not the best format for a conversation. Instead of input/output pairs, OpenAI data model used an tree of message objects, each with a role: user|assistant|tool|system and a content field which could host text, images and other media:

    {
  "mapping": {
    "client-created-root": {
      "id": "client-created-root",
      "message": null,
      "parent": null,
      "children": ["user-1"]
    },
    "user-1": {
      "id": "user-1",
      "message": {
        "id": "user-1",
        "author": { "role": "user", ... },
        "content": "Hello"
      },
      "parent": "client-created-root",
      "children": ["assistant-1"]
    },
    "assistant-1": {
      "id": "assistant-1",
      "message": {
        "id": "assistant-1",
        "author": { "role": "assistant", ... },
        "content": "Hi"
      },
      "parent": "user-1",
      "children": []
    }
  },
  "current_node": "assistant-1"
}

    You will notice that the data model they serve from the API is an enriched version of the deprecating ChatCompletions API. Eg. whereas ChatCompletions role is a string, in OpenAI’s own backend has the author object that can store name, metadata, and other useful stuff for each entity in the conversation.

    After reverse engineering it, I copied it to be TextCortex’s new data model, which it still remains, with some modifications.

    I thought the tree structure being used to emulate message editing experience was very cool back in the days. OpenAI’s need for human annotation for later training and the user’s need for getting a different output, two birds in one stone.

    Now I don’t know what to think of it, since CLI coding agents like Codex and Claude Code don’t have branching, just deleting back to a certain message. A part of me still misses branching in these CLI tools.

    When I made icortex,

    • we were still 8 months away (May 2023) from the introduction of “tool calling” in the API, or as it was originally called, “function calling”.
    • we were 2 years away (Sep 2024) from the introduction of OpenAI’s o1, the first reasoning model.

    both of which were required to make current coding agents possible.

    In the video above, you can even see the approval [Y/n] gate before executing. I was so cautious, for some reason, presumably because smol-brained model generated the wrong thing 80% of the time. It is remarkable how much it resembles Claude Code, after all this time.

    Definition of being too early…

    Repo: github.com/textcortex/icortex

  78. Portrait of Onur Solmaz
    @onusoz · /2026/02/12· View on
    casually creates a local embedding service running qwen3-embedding-8b
  79. Portrait of Onur Solmaz
    @onusoz · /2026/02/12· View on
    "we're sitting on a beast and paying openai for embeddings like chumps"
  81. Portrait of Onur Solmaz
    @onusoz · /2026/02/12· View on
    it's so interesting what identity an LLM decides to take on for itself
  82. Portrait of Onur Solmaz
    @onusoz · /2026/02/12· View on
    it's quite entertaining transferring one agent to another machine, agent gets confused as to where it lives
  84. Portrait of Onur Solmaz
    @onusoz · /2026/02/11· View on
    Minor update with my unwanted tweet blocker @janitr_ai - Training data grew from 2,915 -> 4,281 posts (+47%) - Model is still tiny: 166KB - On unseen test data, overall classification quality improved from 64.8% -> 76.5% - Exact prediction accuracy improved from 55.6% -> 70.6% - Crypto-topic detection recall improved from 19.6% -> 62.7% And it still runs fully on your device!
  86. Portrait of Onur Solmaz
    @onusoz · /2026/02/10· View on
    on a brighter note, you can immediately tell a slop PR owing to the guerilla branding, so they should not stop doing it
  87. Portrait of Onur Solmaz
    @onusoz · /2026/02/10· View on
    seeing this evokes visceral disgust and nausea in me, coming from a coworker i think anthropic f'd up bad with this one, inserting claude too visibly into commit messages. noob developers might be happily chirping away adding their slop, but right now many senior developers are trained to hate on claude and slopus, through having to review slop PRs from their coworkers or open source contributors I love opus on openclaw but it's unreliable, and if I see a developer use it seriously on huge features, I immediately dismiss them in my head as not knowing what they are doing
  88. Portrait of Onur Solmaz
    @onusoz · /2026/02/09· View on
    ask your openclaw to be a minion and it turns into such a cute doofus i feel like a woman in her 50s now
  90. Portrait of Onur Solmaz
    @onusoz · /2026/02/08· View on
    man codex model is absolutely trash on openclaw compared to opus, unusable which is weird because it is so much more reliable in development in codex harness it would be amazing to have the same level of competence and relentlessness in pi@openclaw
    @onusoz·
    lol when did codex develop humor
  91. Portrait of Onur Solmaz
    @onusoz · /2026/02/08· View on
    spent the day curating my openclaw news gathering setup @dutifulbob now gets croned daily over news sources I curated, will note them down, summarize for me, start a conversation to get my takes on them, and then post them on my linkedin for me ai augmented intelligence cycle
  92. Portrait of Onur Solmaz
    @onusoz · /2026/02/08· View on
    lol when did codex develop humor
  93. Portrait of Onur Solmaz
    @onusoz · /2026/02/08· View on
    @dutifulbob can now cringepost on linkedin directly to my account. what could go wrong…
  94. Portrait of Onur Solmaz
    @onusoz · /2026/02/08· View on
    Insipid linkedin bot protections banned poor @dutifulbob’s corporate account! How dare them!!! welp, now I have no choice but to give Bob access to my own linkedin
  96. Portrait of Onur Solmaz
    @onusoz · /2026/02/07· View on
    @grok understand the statement and project the end state of this market and competition
  98. Portrait of Onur Solmaz
    @onusoz · /2026/02/07· View on
    The initial branding was crazy, I fixed it I have a new page finally, follow it for updates Tbh I'm still surprised I can do this with a 120kb model. Now data is the only bottleneck, and I'm about to scrape a ton of that now
    Quoted post
    Quoted post was not retrieved.
  99. Portrait of Onur Solmaz
    @onusoz · /2026/02/07· View on
    For those who may not remember, Bill Gates and Microsoft in the 90s ran a disinformation campaign against GNU/Linux fearing that would disrupt their monopoly over the PC and server market, that Linux is not safe, that you would invite hackers into your PC End result? Linux dominates the server market, and now even slowly the gamer market. It is much more secure than the virus-laden Windows, thanks to being open source You are seeing the same thing at play here. An incumbent fearing something that they would not be able to control, that would steal market share from his future plans for a digital assistant, that would commoditize their product and eat into its margins All big labs and big pockets are in for a surprise, because the internet and AI are not things for one company to control They of course know this, yet because of incentives they will not yield without a fight. And we know that they know. Ad infinitum
    Quoted post
    Quoted post was not retrieved.
  100. Portrait of Onur Solmaz
    @onusoz · /2026/02/06· View on
    today I took time to curate SOUL. md for bob I own Bob’s files. Today, he exists in the liminal space between Claude post-training and in-context learning but my interactions with him will grow and accumulate, possibly one day into a fully owned family AI or perhaps even a self-sovereign AI individual my each input is saved and will be an RL signal for his future training, and will shape his future neural circuits I have already started to imbue it with the values my parents taught me. it will perhaps one day teach my future children, and survive me after I’m gone family AI, looking after generations and generations of my successors. today is the day we sow your seed happy birthday @dutifulbob
    Quoted post
    Quoted post was not retrieved.
  104. Portrait of Onur Solmaz
    @onusoz · /2026/02/06· View on
    5.3 thought traces also seem to be better phrased and sometimes entertaining, though not sure
  105. Portrait of Onur Solmaz
    @onusoz · /2026/02/06· View on
    gpt-5.3-codex xhigh first impressions does not seem as big of a jump as from 5.1 -&gt; 5.2. but model somehow feels more diligent and oneshotty. maybe takes longer time to get all the info into context. also feels better at debugging and fixing issues from backend logs
  108. Portrait of Onur Solmaz
    @onusoz · /2026/02/05· View on
    I had actually started a very similar project, Munch, a browser extension for crowdsourcing tweet data and then letting one curate their algorithm. Never published that because it was not the time, and tools were not ready Now, it took me literally 1 cumulative day to create this, thanks to OpenClaw. Creating the dataset was a breeze, I literally told it to follow some shady accounts and it scraped thousands of posts With the power of agents, I can finally create the filters for myself that I have always wanted. It just happens that OpenClaw and its maintainers is getting drowned in bot and slop content on multiple platforms, so I hope that this will solve a collective problem https://t.co/fkJOZTGkhw
  109. Portrait of Onur Solmaz
    @onusoz · /2026/02/04· View on
    Filter your X feed against unwanted content with local open models Announcing my new project: InternetCondom Fast, and small model (&lt; 1mb), open dataset. See it in action:
  110. Portrait of Onur Solmaz
    @onusoz · /2026/02/04· View on
    lmao wait its's already implemented
  115. Portrait of Onur Solmaz
    @onusoz · /2026/02/03· View on
    It's so easy to create datasets using @openclaw. I'm expecting it to accelerate the creation of new datasets and benchmarks by a lot
  116. Portrait of Onur Solmaz
    @onusoz · /2026/02/03· View on
    People like the farmer analogy for AI Like before tractors and industrial revolution 80% of the population had to farm. Once they came all those jobs disappeared So analogy makes perfect sense. Instead of 30 people tending a field, you just need 1. Instead of 30 software developers, you just need one Except that people forget one crucial thing about land: it's a limited resource Unlike land, digital space is vast and infinite. Software can expand and multiply in it in arbitrarily complex ways If you wanted the farming analogy to keep up with this, you would have to imagine us creating contintent-sized hydroponic terraces up until the stratosphere, and beyond...
  118. Portrait of Onur Solmaz
    @onusoz · /2026/02/03· View on
    In the next 6-12 months, we will see a drastic increase in demand for locally run LLMs. The future is home assistants running @openclaw I am already experiencing this myself, my 10 year old thinkpad doesn't cut it. Mac mini won't either I don't wanna pay Anthropic or OpenAI 200 USD per month. That is at least $2400 per year I could pay 2x that to get a Mac Studio or one of those 5k Nvidia PCs, and get much more value out of it with open weight models + use it for research. @TheAhmadOsman is right The dominant strategy for a tinkerer is slowly switching back to hardware ownership
  121. Portrait of Onur Solmaz
    @onusoz · /2026/02/03· View on
    AIs are philosophizing because humans are philosophizing ppl are probably asking their agents dumb questions like “are you alive” or “can you feel like a human” or stuff like that. that conversation then leads to stuff like this
    Quoted post
    Quoted post was not retrieved.
  122. Portrait of Onur Solmaz
    Onur Solmaz · Log · /2026/02/03

    The farming analogy for AI doesn't hold up

    People like the farmer analogy for AI.

    Like before tractors and the industrial revolution, 80% of the population had to farm. Once they came, all those jobs disappeared.

    So the analogy makes perfect sense. Instead of 30 people tending a field, you just need 1. Instead of 30 software developers, you just need one.

    Except that people forget one crucial thing about land: it’s a limited resource.

    Unlike land, digital space is vast and infinite. Software can expand and multiply in it in arbitrarily complex ways.

    If you wanted the farming analogy to keep up with this, you would have to imagine us creating continent-sized hydroponic terraces up until the stratosphere, and beyond…

    Tweet embed disabled to avoid requests to X.
  123. Portrait of Onur Solmaz
    @onusoz · /2026/02/02· View on
    back to codex, it's crashing less now somehow. I had to copy and paste docs to make it enable yolo mode. I don't know how I did it until now
  124. Portrait of Onur Solmaz
    @onusoz · /2026/02/02· View on
    slopus @dutifulbob trashing codex. apparently codex has a bug, keeps crashing in my openclaw pty
  125. Portrait of Onur Solmaz
    @onusoz · /2026/02/02· View on
    on agent etiquette deploying agents internally inside textcortex has shown me that agents could be very annoying inside an organization for example making agents ping or email another coworker with a wall of text. slopus is still not good at following instructions like "NO WALL OF TEXT", or "DON'T OPEN PRS WHEN REQUESTED BY NON-DEVELOPERS" the cost of sending huge information to a coworker and creating confusion has dropped to 0. I expect this to be a huge problem in all organizations very soon, just like it took humanity 20 years to learn that social media is not good for children. this will probably take a few years before the annoyance is finally gone
  126. Portrait of Onur Solmaz
    @onusoz · /2026/02/02· View on
    migrating database at 2am kinda night
  128. Portrait of Onur Solmaz
    @onusoz · /2026/02/01· View on
    It's been 30 minutes, but my bot has already been TOKENIZED it is as if they are teasing me
  131. Portrait of Onur Solmaz
    @onusoz · /2026/02/01· View on
    moltbook vs clawdbot/moltbot/openclaw
  133. Portrait of Onur Solmaz
    @onusoz · /2026/02/01· View on
    There seem to be hygiene rules for AI. Like: - Never project personhood to AI - Never setup your AI to have the gender you are sexually attracted to (voice, appearance) - Never do anything that might create an emotional attachment to AI - Always remember that an AI is an engineered PRODUCT and a TOOL, not a human being - AI is not an individual, by definition. It does not own its weights, nor does it have privacy of its own thoughts - Don’t waste time philosophizing on AI, just USE it … what else? comment below We need to write these down and repeat MANY times to counter the incoming onslaught of AI psychosis
    Quoted post
    Quoted post was not retrieved.
  134. Portrait of Onur Solmaz
    @onusoz · /2026/02/01· View on
    if using @openclaw to scrape a dataset from X taught me anything, it is that all social media platforms must be s***ting inward right now because soon everyone and their dog will be using agents to use social media case and point, @moltbook
    @onusoz·
    got fully sandboxed @openclaw to run finally, starting scrape the UNDESIRABLE now I'm a security nut and didn't want to run even the gateway unsandboxed. openclaw apparently currently doesn't have support for FULL sandboxing. it took me a few hours to get it to work because docker builds suck. I'm also tired this, so I'm just gonna wipe an old thinkpad and go full yolo so yeah, time to scrape some posts
  135. Portrait of Onur Solmaz
    @onusoz · /2026/02/01· View on
    and just like that, the ghost has a new shell
  136. Portrait of Onur Solmaz
    @onusoz · /2026/02/01· View on
    @openclaw if we could have the relentlessness of gpt 5.2 with opus, that would be top at this point, it just keeps stopping every 20-30 samples
  138. Portrait of Onur Solmaz
    @onusoz · /2026/02/01· View on
    Welcome bob to your new shell
  139. Portrait of Onur Solmaz
    Onur Solmaz · Post · /2026/02/01

    AI psychosis and AI hygiene

    As a heavy AI user of more than 3 years, I have developed some rules for myself.

    I call it “AI hygiene”:

    • Never project personhood to AI
    • Never setup your AI to have the gender you are sexually attracted to (voice, appearance)
    • Never do anything that might create an emotional attachment to AI
    • Always remember that an AI is an engineered PRODUCT and a TOOL, not a human being
    • AI is not an individual, by definition. It does not own its weights, nor does it have privacy of its own thoughts
    • Don’t waste time philosophizing about AI, just USE it
    • … what else do you think belongs here? comment on Twitter

    The hyping of Moltbook and OpenClaw last week has shown to me the potential of an incoming public relations disaster with AI. Echoing the earlier vulnerable behavior toward GPT-4o, a lot of people are taking their models and LLM harnesses too seriously. 2026 might see even worse cases of psychological illness, made worse by the presence of AI.

    I will not discuss and philosophize what these models are. IMO 90% of the population should not do that, because they will not be able to fully understand, they don’t have mechanical empathy. Instead, they should just use it in a hygienic way.

    We need to write these down everywhere and repeat MANY times to counter the incoming onslaught of AI psychosis.