32 comments

  • mariocesar5 hours ago
    I'm was closing a day of work and it stopped in the middle of a long plan, I started codex and tell it: "I was in the middle of work with Claude could you read the plan and the session "x" in ~/.claude and continue the work" it just completed everything :D
    • jerpint4 hours ago
      I’m working on a project that makes switching between coding harnesses essentially unnoticeable. It’s particularly useful when I run out of credits on any given day<p>The entire platform is skill driven, and based on the premise that state is your local file system. That makes switching harnesses so easy<p>It’s all open source and has plenty of other features, including inter agent communication, telegram client and much more in the pipeline<p><a href="https:&#x2F;&#x2F;www.woltspace.com&#x2F;" rel="nofollow">https:&#x2F;&#x2F;www.woltspace.com&#x2F;</a>
      • calgoo1 hour ago
        Not sure what the problem with switching is? When I run out of my kimi session, I just switch model and tell deepseek to continue. Yes I eat the initial cache miss but that&#x27;s it, no fancy harness required.
      • etoxin2 hours ago
        I think having a codebase that can use any harness is the optimal setup. I often switch between harnesses and models all the time at work. We have our state&#x2F;spec stored in git, so we can pack it up on friday and start fresh on monday. Or more commonly, using gemini&#x2F;opus to create rich specifications and using Luna to implement them. Then switching again for reviews etc.
      • vasco22 minutes ago
        No project needed man. There&#x27;s also a whole ycombinator company for the same thing, skillsync, also useless. These things are either out of the box or they are 2 prompts away.
    • ammario2 hours ago
      Related, have found this handoff skill invaluable for moving tasks between harnesses: <a href="https:&#x2F;&#x2F;gist.github.com&#x2F;ammario&#x2F;cd57a1d7f6c911e079378c540b384c4a" rel="nofollow">https:&#x2F;&#x2F;gist.github.com&#x2F;ammario&#x2F;cd57a1d7f6c911e079378c540b38...</a>
    • RGS18115 hours ago
      I did this with DeepSeek a couple of days ago.
      • oulu20065 hours ago
        yeah same that&#x27;s my preferred flow now -- sill use Claude and OpenAI a bit, though recent nerfs to subs has really made using codex much harder.<p>DS 4.1 flash is my main powerhouse and Opus&#x2F;Astra my auditors (when they&#x27;re not out of tokens) otherwise K3 or DS4 pro
        • SOLAR_FIELDS5 hours ago
          I started using Pi with Astra on a whim after really enjoying Astra and reading somewhere that you get close to identical results as with the codex harness but for a significant hunk less token usage.<p>Coming from mostly using Claude models, the terse factual statements coming from Astra via the Pi harness are a breath of fresh air over having to wade through the flowery verbose nonsense that Claude constantly outputs
          • pdntspa1 hour ago
            I recently had Astra review a fairly detailed design doc I have for an audio VST fork, that I originally wrote with Opus and&#x2F;or Fable a few months ago. The doc reaches deep into signal flow and module topology while lifting most of the DSP code from other open-source projects. I had it review for feasibility and architectural soundness.<p>Astra found a number of flaws that would have come up during implementation and we worked through them. But then I had Fable 5.1 review <i>that</i> document and it found a number of issues with Astra&#x27;s changes, the least of which had was that Astra duplicated a lot of technical notions that it added rather than using references to an authoritative section. It also flagged some of Astra&#x27;s designs as technically impossible, pointing out why and I&#x27;m actually in the process of digesting its feedback and updating the design spec. (I hand-review each point and we work through a solution together -- I don&#x27;t trust either model to come up with something that follows my vision on their own)<p>I&#x27;m not promoting one or the other, I just found it interesting how this sort of adversarial review found pretty significant flaws in the other model&#x27;s work. I am curious as to whether this process will eventually converge on a document that both agree on or if the models are going to perpetually nitpick each other.<p>I haven&#x27;t actually started implementation yet, so maybe one or the other is full of shit. Just trying to come up with an architecturally sound design for something I want to write, when I lack the DSP knowledge to be able to write it myself. But the intent is to pass an agent the design doc and list of milestones and let it handle implementation.
        • consumer4515 hours ago
          Ignorant questions for you and anyone else using DS:<p>1. who hosts the inference<p>2. which harness are you using with it, still CC?
          • oulu20065 hours ago
            So everyone has their preferred way of doing it.<p>1. I go direct to source, i.e. DS platform, I find it cheaper than paying the openrouter tax -- I also switch it up a bit<p>2. I built a local LLM router, that I update with new profiles that have my preferred provider of the week (lowest token costs&#x2F;speed) with fallbacks, like mimo --&gt; DS4 etc.. if there is overloading,<p>3. I use 3 diff harnesses, CC&#x2F;Codex + Opencode -- they all talk to each other through a custom rig system that routes messages between llms using a Rust backed structured JSON system<p>Not saying this is the best, it&#x27;s just what I like and works for me^.<p>I can flow quite naturally between Opus&#x2F;Astra&#x2F;K3&#x2F;GLM&#x2F;MiMo&#x2F;DS&#x2F;etc.. this way and often do...more so these days with subs no longer great as they used to be.
          • logicchains26 minutes ago
            Deepseek harness is great!
          • pavo-etc5 hours ago
            Whatever openrouter puts me on, running in pi (though I do all comms over my xmpp wrapper).
    • ebbi5 hours ago
      how do you get the &quot;x&quot; session name?
      • drewnick4 hours ago
        You can even tell the new harness to go find your latest session in the other harness and it works perfectly. You don&#x27;t even need the name!
      • consumer4512 hours ago
        In CC, &quot;&#x2F;exit&quot; will tell you how to <i>--resume &lt;session-id&gt;</i> upon exit. This will eat zero tokens, and it actually does not require an active internet connection.
      • mariocesar4 hours ago
        `claude --resume` will list your sessions, you can copy the title. Codex may just list and grep the sessions files in ~&#x2F;.claude and filter the one with the name.<p>I also have the habit of naming my sessions with `&#x2F;rename`
      • simlevesque5 hours ago
        &#x2F;resume in claude code.
  • prodigycorp5 hours ago
    Opus 5.5 and fable 5.2 will release tonight.<p>gpt-6-sol and Aeon (personal agent) on Thursday. Already preceded by a huge week with step, mimo, grok, and jev releases.<p>Relentless cycle.
    • Razengan4 hours ago
      In a way it&#x27;s like the 1980s again, when there were new computers and consoles coming out every week, or the 2000s with all the 3D cards :)
      • shepherdjerred3 hours ago
        Wow was the 80s really like this? I am interested in AI but it is a bit exhausting to keep up with
        • Baeocystin46 minutes ago
          It&#x27;s hard to overstate how much computers improved with each generation back then. Going from, say, an IBM XT or AT to an Amiga has no modern parallel. It was mind-blowing, exhausting, and exhilarating all at once.
          • Razengan1 minute ago
            You could literally count the number of extra colors you would get onscreen with each new computer!
      • prodigycorp4 hours ago
        Whoa, no kidding. Takes me back to the MHz wars we had at school and like you said, the card wars. Excellent perspective.
        • system24 hours ago
          At least after buying 166mmx you didn&#x27;t get lowered to 133mhz after a month. The progress was legit back then.
          • AnotherGoodName4 hours ago
            Hey I bought a ‘cyrix pr166’ so i’m not too sure about that.
  • tombert4 hours ago
    I don&#x27;t know if something changed recently, but I have been getting a <i>lot</i> of stuff &quot;flagged by safeguards&quot; in the last two days.<p>I am quite confident that what I&#x27;m doing is well within the law, and I&#x27;m not even doing any kind of pen-testing stuff, just some basic reverse engineering, but I can&#x27;t even use Fable anymore because every time I enable it, it works for about twenty seconds and makes me drop down to Opus 4.8, and often even down to Sonnet.<p>If anyone here works at Anthropic, did you make the safeguards super sensitive recently?
  • theophilus765 hours ago
    Claude is both on the verge of replacing all software devs and keeping a two 9 SLA
    • chrisdbanks5 hours ago
      Your average dev is sick 7 days a year so it&#x27;s only natural
      • blitzar2 hours ago
        10x dev ... so sick 70 days a year is the target
      • Hamuko2 hours ago
        I wonder if I&#x27;m being saved by remote working since I think I have three days in the 365 days, and I&#x27;ve also managed to go years without a single sick day.
    • nikcub4 hours ago
      and the two 9&#x27;s aren&#x27;t in order
    • Schiendelman5 hours ago
      Sure, software engineers have less than two nines. :)
  • prologic5 hours ago
    Good timing too, Just as xiaomi[1] makes their big announcement :D<p>[1]: <a href="https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=49792730">https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=49792730</a>
  • joegibbs6 hours ago
    Also for Grok at the same time, must be a problem with Colossus
  • camkego5 hours ago
    When this happens, how often does it cause us to lose our cached pricing? (when we should be getting the cached price)
  • makeavish6 hours ago
    There are rumors of Opus 5.5 launching on Tuesday so maybe it due to that? Previously as well there have been incidents just before launch.
    • Wowfunhappy5 hours ago
      Do you know where these rumors are coming from? I keep seeing posts like yours on HN and Reddit, which, yes, literally confirms there are rumors, but I can&#x27;t tell whether they&#x27;re based on anything.
      • simlevesque4 hours ago
        All their API routes for non existing models return 404 and when there&#x27;s a new model it returns 403.
  • homo__sapiens5 hours ago
    Millions of people thought process disrupted.
  • NoPicklez1 hour ago
    You&#x27;re telling me I have to switch to using my brains tokens now?
  • bonsai_spool5 hours ago
    Posted this in the xiaomi thread, but are folks working in biology &#x2F; cybersecurity seeing more limits in what Opus (not Fable) is allowing? This has happened quite suddenly for me and I’m stuck in the middle of a project that would have otherwise called for use of Claude.<p>I’ll be trying these models out and may end up switching my subscriptions if this craziness continues
    • bbeonx5 hours ago
      i was talking to opus the other night about the polio vaccine related stuff (i had been arguing with one of those people who claims the polio vaccine didn&#x27;t work so i figured i&#x27;d use it as an excuse to brush up on some bio&#x2F;history). i asked it to explain how polio virus was isolated pre PCR and claude hard killed my session citing security reasons (it was talking about stool samples ffs LOL)
      • AnotherGoodName4 hours ago
        Hell i’m working on ai for board games as a hobby (tfmbot.com) and one of the cards is ‘microbes’. Sent straight back to 4.8 for having that in my code.
      • a0124 hours ago
        In Anthropic excuse, stool samples is bio weapon if you tried too hard
  • mattdeboard6 hours ago
    Their &quot;Subscribe to incident&quot; doodad isn&#x27;t working either. Hanging when posting contact info object
    • porlex6 hours ago
      It seems to be updated correctly now
  • jcims4 hours ago
    Wonder if it had anything to do with this - <a href="https:&#x2F;&#x2F;health.aws.amazon.com&#x2F;health&#x2F;status" rel="nofollow">https:&#x2F;&#x2F;health.aws.amazon.com&#x2F;health&#x2F;status</a>
  • ehnto4 hours ago
    Apologies for the rant, but why do these things constantly hit the front page? It is not interesting, it&#x27;s not a discussion, and if you are using the models you probably already know.<p>HN is already a waterfall of AI meta conversations and bike-shedding, now we have to discuss service outages about the AI too?<p>Can we talk about stuff people are building again, with or without AI, and stop gasping at every minute detail of LLM service providers.
    • nullc3 hours ago
      Gamblers talking about their favorite casino.
    • system24 hours ago
      Everyone is collectively hating on Anthropic, so this news is just adding to the fun.
  • bombcar5 hours ago
    &#x2F;model claude-opus-4-8<p>and keep workin&#x27;
    • TYPE_FASTER5 hours ago
      Yup, switched back to good ol&#x27; opus-4-6, seems to be working for now.
    • bdangubic5 hours ago
      like it&#x27;s 1999 :)
  • thenipper6 hours ago
    us-east-1 was blowing up a bit tonight with ec-2. Maybe it&#x27;s related.
  • yrcyrc5 hours ago
    Was in the middle of a server firewall change and reboot, killed the kernel, booted me out. Had to sweat the recovery a bit.
  • throw567643u86 hours ago
    Something is on fire. Literally or figuratively.
    • windexh8er5 hours ago
      Also... One of the only industries&#x2F;products where unusable product&#x2F;seevice (token &#x2F; context burned) is neither refunded let alone even acknowledged. I wonder how many millions of dollars of inference have been stolen during these outages? First you lose the work in flight, then you lose the time, and finally you lose the token burn for a second time. What a fantastic time to not have any regulations if you&#x27;re rich and greedy!
      • Schiendelman5 hours ago
        Just like a software engineer who gets sick or has to pick up their kids or has an appointment. We&#x27;re used to this, oddly.
  • guybedo6 hours ago
    problems with Grok too ... so i guess Grok 4.7 tried to escape and took control of Colossus datacenter.
  • joshtronic5 hours ago
    When something happens regularly, is it really front page of HN worthy?
  • rvz6 hours ago
    Everyone at Anthropic is interviewing a new Claude model which they are about to hire and forgot to monitor their old model and thus it goes down.<p>If outages like this happened on every major deploy at any other company; this would be viewed as unacceptable, especially if it was something like Google Search going down on every update.
  • krembo2 hours ago
    I have a deja vu like this happens everytime before they launch a new family of models
  • bofia3 hours ago
    What’s the SLA?
  • calvinmorrison6 hours ago
    &gt; Current Status Powered by Atlassian Statuspage<p>Proof that vibe coded or not, people pay someone else mostly for liability.
    • corvad6 hours ago
      I think thats more for being able to have the status page not tied to your own infrastructure.
      • Aeolun5 hours ago
        It’s super easy to just run a totally disconnected synthetic check.
    • bugfix6 hours ago
      If you host it yourself, an outage on your side might end up taking it offline too.
    • 151556 hours ago
      And uniformity.
  • acedTrex5 hours ago
    Excellent news everyone, coding is solved! you can see the results for yourself.
  • dougame2 hours ago
    [flagged]
  • dragonlin5 hours ago
    [flagged]
  • loverofspades3 hours ago
    [flagged]
  • iammrpayments5 hours ago
    I’ve subscribed to chatgpt at 50% off last week and I’m using claude less and less. Not only replies are faster but 20$ plan gives you access to astra.<p>Not sure how can claude even compete at this point unless openAI seriously downgrade the models to save money.
    • moecables5 hours ago
      how&#x27;d you get the 50% off?