... real provider and at our sandbox and diffs the result. It's already caught four real drifts in our own simulation. @vansh36 flagged the favicon was still the Next.js default. Fixed. Half done: @maschiojv , @hi_i_am_mimo , @omri_ben_shoham1 and @priya_kushwaha1 , out-of-order events. The permutation fuzzer is built and Paddle has the patterns. What isn't wired yet is the end-to-end path where you describe the ordering bug and it reproduces ... ... demand. Not going to say it's done when I just checked and it isn't. Not done : @sabber_ahamed , per-event latency jitter. @omri_ben

... rather than from a model impersonating one. Which is maybe the honest version of where we are: the artifact is measurable and the intent is not, so we are all measuring the artifact and calling it intent. Comment from Ben Cohen(@bencohen): @serdarships   @niveditha_patluri1  That s actually really interesting. And yeah, treating it as a floor rather than a score makes a lot more sense. 0 vs

... write down why. Not because Cloudflare's defaults are wrong, mostly they're reasonable, but because a default you never chose becomes a strategy by attrition. This one has a date on it. Comment from Serdar D(@serdarships): @omri_ben_shoham1   Ran it. Real ID with a known current status, interleaved with the invented one in the same batch, four repeats each, same spacing. Result: no disagreement anywhere. The real ID returned an identical name and price ... ... only detect instability, and a deterministic mislabel passes it every time, by construction. The control arm is what exposed it, and only because the two arms failed in shapes that should not have matched each other. Comment from Omri Ben

... solution that could actually solve my problem, I ll give it a try. Although, we can never know for 100% if upfront setup will actually deliver the automation we need. Comment from Adana Marukhyan(@adana): @omri_ben_shoham1  Absolutely agree with your last sentence. And I agree that products that unify the interface rather than the backend are becoming a very interesting modern approach. Do you use any ? Comment from Philip Roigard(@philip_roigard): @omri_ben

... logsheet.ch and the app is LogSheet on the App Store. Happy to put whatever numbers I have in the comments including the unflattering ones. If you want to see the app and leave a score! apps.apple.com/app/id6795027347 Thanks Ben ...

... this is universal. what's a feature that you almost didn't ship, or built half-heartedly, that ended up mattering way more than the stuff you were actually proud of building? Comment from Venkat A(@venkat_a2): @omri_ben_shoham1 Mine is a feature that refuses to do the thing the product is sold on. We turn a customer interview into a written case study. Somewhere in there I added a check that judges whether the conversation actually produced ... ... sound the same. Comment from Evelyn Clark(@evelyn_clark1): Cool story. I'd bet it's user-base dependent too different crowd here, technical folks lean into raw/debug tools more than a general audience would. Comment from Omri Ben

... happened, they just inherited a "trusted" input. curious if anyone building for the tool-calling case has a version of the destination check for a consumer they don't operate and can't instrument. Comment from Omri Ben-Shoham(@omri_ben

... everyone, I m Ben. I run two businesses, and somewhere between meetings, driving, client work, and everyday life, I m constantly having ideas or remembering things I need to do. Writing something down was never really the problem. The problem was what happened afterward. My notes ended up scattered across different apps, transcripts, documents, and random lists. Even when an AI tool cleaned up or summarized a recording, I was usually left with another large block of text that ... ... under my launched apps I m happy to answer questions about the product, how I built it, or what I ve learned trying to launch it while running two other businesses. Comment from Noah Zaleman(@noah_zaleman): @ben

... Coding Agent , and Hermes. I m curious. Which agent harnesses are you currently using? How do you use them differently? Am I missing any that you would strongly recommend? Looking forward to hearing everyone s thoughts! Comment from Omri Ben-Shoham(@omri_ben

... find them first. we now throw every ordering + every duplicate at your handler and hand back the exact sequence that broke it. which of these has bitten you? (i've shipped #1 myself. be honest ...) Comment from Omri Ben-Shoham(@omri_ben

... roadmap for august.., not shipped. so today it's the spec + the invariants doing the heavy lifting, and I watch the drift piece by hand. rather tell you that than pretend it's automated. Comment from Omri Ben-Shoham(@omri_ben

... have a page, no? I think that people should have a landing page for their product to describe more precisely the product + prelink it :) Comment from Sabber Ahamed(@sabber_ahamed): @busmark_w_nika   @abdullah_javaid3  the invitation vs favor framing is the most useful line in this whole thread. i caught myself drafting DMs that were basically upvote asks in a nicer outfit. what actually worked was showing up in threads like this one for weeks with nothing ... ... nika  Fair point, and yeah I overcomplicated it Simplest version: run more than one product, and most CRMs turn into a mess. LiftUp keeps each product's stuff cleanly separate so you always know what's what. Competitors

... Right now it often adds to it more tasks in flight, more context to hold, more decisions to make. Curious if you actually feel it as freeing, or if the discipline is what makes it work. Comment from Omri Ben-Shoham(@omri_ben

... covered, the downside basically disappears. If it's not, "ugly but working" beats "pretty but unverified" every time. So it's less about the code and more about how much I trust my tests around it. Comment from Omri Ben-Shoham(@omri_ben

... sure, now, I am asking already happy users to rate my product as best as possible :D Because harsh times can struct and I need to be prepared :D Comment from Nika(@busmark_w_nika): @omri_ben_shoham1  That's why I am trying to ask satisfied users for reviewing my solution :D Comment from Nika(@busmark_w_nika): @marius_stoicescu  Sometimes it is better to skip our native countries. I am from Slovakia ... ... delete them :D Comment from Nika(@busmark_w_nika): @richard_romero2  I have heard one good advice. When people are too harsh, just close the laptop for a while :D Comment from Nika(@busmark_w_nika): @omri_ben

... recurring is the part I'd want to stress-test most in practice, curious how tight that gate feels once you're a few weeks into real usage rather than a demo. Comment from Omar Ghandour(@omarships): @omri_ben_shoham1  Neither, actually, and that's kind of the point. A conflict doesn't fire off a ticket to someone. Both facts sit as competing edges and wait, and most resolve on their own: new data flows in and confirms ... ... through it, the more the brain holds, and your agents draw on all of it to actually do the work. It's not waiting to learn you, it starts with your context and compounds from there. Comment from Omri Ben

... this side project is teaching me that you don t need to understand your own codebase deeply to build something people love. You just need to understand the problem. Anyone else navigating this weird middle ground? Comment from Omri Ben-Shoham(@omri_ben

... that would be expensive to re-litigate. Feeding that back in beats re-explaining from scratch, though it only works if I actually update it right after a decision instead of trying to reconstruct it later. Comment from Omri Ben-Shoham(@omri_ben

... voice, but ultimately, won't let it hit "send". It too often doesn't quite get it! But I also have a consumer health app, so more sensitive in this domain. Comment from Sabber Ahamed(@sabber_ahamed): @omri_ben_shoham1  same silent-failure logic is why anything writing to a health record gets a human pass here too - a wrong value in a checkin log doesn't throw, it just sits there quietly wrong until someone acts ... ... apart. In general, I think it's really important to build a system where the human in the loop shows up in the right, necessary, and important places, instead of handing everything over to the agent. Comment from Omri Ben

... anyone gotten pushback for shipping code they couldn't fully explain when something went wrong? - Has this changed how thoroughly you review AI output before it ships, or has trust just gone up over time? Comment from Omri Ben-Shoham(@omri_ben

... what it actually does: It sees your screen and acts on it Not summarizes. Not suggests. Clicks, scrolls, fills forms, and navigates any page, on its own. It builds Google Sheets and Docs from scratch "Make me a competitor pricing table with graphs." It opens the tabs, pulls the data, formats the sheet, builds the charts. You just watch. It runs WhatsApp & Telegram like a personal assistant Broadcast to 50 contacts at once. Send follow-ups with Word docs ... ... attached. Message from your phone, get a finished file back. Nobody else does this. It monitors pages and pushes you alerts Price dropped? Ticket available? Competitor

... were an experienced programmer, I would just ask AI where the particular parts in the code are responsible for the feature, and I could change them later. Once the man has the knowledge, everything is easier :) Comment from Omri Ben-Shoham(@omri_ben

... working on exposing an eval environment configuration. The use cases we discovered The ones we expected: simulation for learners, architecture validation, cost estimation. The ones the community surfaced: Architecture reviews before anyone writes an infrastructure ticket. Omri Ben-Shoham @omri_ben

... reporting script. That was the moment I realized scaling isn't just about whether the AI can solve the problem it's also about whether the workflow stays efficient as the project grows. Comment from Artem Tuleninov(@redist): @omri_ben_shoham1  Exactly "never trust the summary, always check the system of record" is probably the most underrated rule in agent development right now. The transcript is the agent's story about what happened. The calendar/database/CRM is what actually happened ... ... Those two things can diverge silently and confidently at the same time. Building that verification layer in from the start changes how you architect the whole thing. Comment from Omri Ben

... worked in the demo, then it fell over" stories in there that are way more useful than the marketing copy. Happy to be here, looking forward to more of these threads. Comment from Omri Ben-Shoham(@omri_ben