... through unredacted or stop? Write the answer down. Third, test the guard by trying to get around it. Feed it the awkward input, kill the workflow midway, and see what the agent does next. For an app like Murror, where people write things they wouldn't say out loud, how a reflection is worded is a judgment call a model is good at. Whether an entry is handled according to the privacy rules we've promised shouldn't depend ...

... disclose and protect. If your product handles anything personal, the question isn't "can we afford to send this?" It's "does this answer get better enough to be worth sending it?" We learned this the hard way at Murror. We used to keep every journal entry forever because more history looked like better insight. When we capped retention at 30 days, the insights barely changed and people started writing more openly. The context we thought we needed was mostly ... ... check is the one you should be doing anyway: do your buttons and fields have real accessible labels? If you're choosing a model this week, what are you measuring besides price per token? Comment from Mona Truong(@monatruong_murror

... Anything that can read untrusted content (email, web pages, support tickets) shouldn't also be able to send, delete or publish without a person approving it. Third, keep sensitive data out of reach by default. For an app like Murror, where people write things in a journal they wouldn't say out loud, the rule I'd apply is that nothing in your own tooling gets a standing grant to that kind of data. If you need to look ... ... October. It's brief, so read it directly instead of relying on my summary. What's the broadest access you've given an agent so far, and would you give it again today? Comment from Mona Truong(@monatruong_murror

... that the age rating questionnaire now includes questions about social media capabilities, and that "beginning in September 2026, responses will be required when submitting new apps or updates to the App Store, or when submitting apps for notarization for alternative distribution." September ends tomorrow, so if you have a build in flight, this is the week you meet it. Apple's definition is narrow: the ability to redistribute, amplify, or interact with user-generated content through a social feed ... ... similar discovery method. If that is not your app, you answer no and move on. Murror

... adjacent to this. You're the definition. I build an app that helps people process their emotions with an AI that remembers what they told it last time. When I went looking for whether that puts Murror inside any of the new AI-companion laws, the honest answer was: probably, and I wouldn't have guessed that a year ago. Three states have live statutes now, not proposals. California's SB 243 was signed October 13, 2025, and its operator ... ... rather than a companion claim. Write down, in one place, what the product actually does when someone's input looks like a crisis - not what I assumed it does. Add the disclosure and the three-hour reminder even though Murror

... live right now. Reporting obligations under Article 14 have applied since 11 September 2026. That's eleven days ago. I went and read the Commission's own reporting page and ENISA's platform FAQ rather than the coverage, because Murror ships a mobile app into the EU and I wanted to know what I'd actually have to do at 2am. Three things in there surprised me, and one of them is a live bug in the platform. First, whether ... ... absence would stop the product doing one of its functions. If your app is a thin client over your API, that API is not a separate service you can wave at. It's part of the product. This product-versus

... nobody covered, which is worse, because it changes a line of code a lot of us shipped this year without thinking about it. If your product has a "paste a URL" box, a link unfurler, an enrichment step, a competitor-monitoring job, or anything that fetches a page and hands it to a model, this is about you. What was actually measured. The paper is "Indirect Prompt Injection in the Wild" by Khodayari, Zhang, Acharya and Pellegrino, out of CISPA ... ... compliance rate on a summariser is an annoyance. The same 1% on something that can send, delete, or pay is a different sentence entirely, and 14.2% of Unit 42's detections were aiming at exactly that. For Murror

... right and is more useful than either, and the revised version went up this month, so it's worth twenty minutes before you write your Q4 plan. It's NBER working paper 35275, Demirer, Musolff and Yang - "Writing Code vs. Shipping Code." More than 500,000 GitHub developers, matched against their actual AI usage telemetry rather than a survey, in a matched event study design. The headline numbers are large and I have no interest in arguing with them. Cumulative ... ... your commit history. If commits are up and releases are flat, write down what sits between them, in order. That list is your Q4 roadmap, and almost nothing on it is a coding problem. I did this for Murror

... line appears in both: starting September 2026, you're required to indicate whether your app or game includes social media capabilities in order to submit new versions or updates to the App Store, or for notarization for distribution on alternative marketplaces. It's September. It's live. I read both pages this week rather than the coverage, partly because Murror

... Apple's own support pages. Apple In-App Purchase: 26%, or 15% if you're in the Small Business Program, the Mini Apps or Video Partner Programs, or on an auto-renewable subscription after the subscriber's first year. Alternative payment processing inside your app: 20%, or 10% for those same groups. Out-of-app offers with an actionable link: a 15% store services commission, or 10% for those groups, on anything bought within 7 days of the tap. Apps ... ... distributed outside the App Store entirely - an alternative

... Three numbers from it get quoted. Users who saw an AI summary clicked a traditional result in 8% of visits; users who didn't, 15%. Users ended their browsing session 26% of the time with an AI summary present versus 16% without. And the one that should matter most to anyone buying GEO: users clicked a link inside the AI summary itself in 1% of visits. One per cent. That is the click value of the thing the entire optimisation ... ... more often. Just 8% of one- or two-word searches produced one. For searches of ten words or more, 53%. For searches starting with a question word - who, what, when, why - 60%. Sit with that next to the 8%-versus

... answer is text, you're outside the emotion recognition definition. If it's voice, face or anything physiological, you're in a different conversation and should have it with a lawyer rather than a forum post. The Murror version is short. We're text-based, so on the definition we aren't an emotion recognition system, and we say we're AI in the first screen because a journal that pretends to be a person is a bad journal. What ... ... times. I'm a founder, not a lawyer, and none of this is legal advice - but the arithmetic of which paragraph applies to you is something you can do yourself in twenty minutes, and it's cheaper than the alternative

... recovered. There's a version of this story going around that says Android is closing. Sideloading is dead, the open platform is over, verify or disappear. I went and read Google's own developer verification pages this week, because Murror ships on Android and I wanted to know what I actually had to do before the end of the month. The docs say something narrower than the panic, and in one specific place something worse. Start with what happens ... ... full-distribution fee on the Android Developer Console, waived for limited accounts. And if you're an organisation, start the D-U-N-S now. Twenty-eight days is the kind of number that eats a deadline quietly. The Murror

... There's a distribution channel a lot of makers are quietly building for right now, and the rulebook for it is a page almost nobody reads. I read it this week because I was working out whether Murror belongs there, and one paragraph changed how I'd plan around it. Here it is, from OpenAI's app submission guidelines rather than the coverage: "Currently, apps may conduct commerce only for physical goods. Selling digital products or services including subscriptions, digital ... ... what you wanted. The docs say "currently" a lot. That word is doing real work, and the version of you who reads the changelog in six months will not reconstruct today's reasoning. Where I am on it. For Murror

... what the app does? Is broadening what the assistant will discuss? I don't think anyone knows yet, and I'd rather have written down my answer before a release forces one. Where I actually am on this. Murror is an emotional-reflection product, so the age of the person on the other side is not a compliance field to me, it's the most load-bearing context I could have. And I still don't have a clean answer ... ... verification-method question, because the honest version is that I don't yet know what I'd do differently with a government-ID-checked 16 versus

... stance about a customer. Those got put on the same settings page next to each other, in the same visual language, and they are not the same decision at all. I don't have a satisfying answer for Murror. We're not ad-supported, so the default doesn't bite. But the honest reason I'd never block Agent is thin: I want to be findable, and right now agents are how a growing share of finding happens. If people ... ... 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 Omri Ben-Shoham(@omri_ben_shoham1): Good writeup - the distinction between blocking Training vs

... that ends, a reflection that concludes reads as thoughtful. A hard cap that fires mid-use tells your most invested user she's used too much of the thing she's paying for. Same economics, completely different product. The Murror version of this is uncomfortable and I'd rather say it than imply I'm above it. Our heaviest users are people going through something. The person journaling every day is in a hard month, and she is simultaneously ... ... payoff shows up a median of 17 hours later, often more than a day. so the heavy cohort reads as the worst margin right up until you widen the window, and then it flips. Comment from Mona Truong(@monatruong_murror

... just name it rather than pretend. I post here because it's where the people I want to talk to actually are, and I'd be writing this the same way if none of it were ever cited. Murror sits badly inside all of this, which is why I've been thinking about it. Nobody searches for an AI companion for understanding your emotions. They type something like "why do I keep having the same fight." The gap between ... ... mature: keep a fixed set of 20-30 non-brand problem prompts, record whether your category and product are mentioned or cited, then separately track actual AI referrals and conversions. Those are three different signals. For a product like Murror

... Board have assessed it as adequate, and adhering to it is the cheapest way to demonstrate compliance on the marking duties. Opting out means demonstrating compliance some other way and fielding more requests for information. The Murror version: we infer emotional state from what people write, and nothing else, so 50(3) isn't ours. 50(1) is, and I don't get to lean on "it's obvious" someone arriving from an ad in a bad week ...

... never sees your brand. Distribution without a front door. But I want to make the harder argument, because I live on the other side of it. Not every product should be a tool an agent picks up. I build Murror, an AI companion for understanding your own emotions, and "expose it as an MCP tool so agents can call it" is a genuinely bad idea for us, and it's worth being precise about why. The value of Murror

... your own evidence otherwise, and price your acquisition spend as though most of it evaporates. And stop reading engagement as a health metric without a second signal next to it. "Sessions down" is now a question, not an answer. Murror sits awkwardly inside all of this, which is why I've been chewing on it. We're a journaling and reflection product. Someone using us less could mean they've drifted off, or it could mean they're doing better ...

... word on anything that matters. A judge with a length bias will happily tell you your more verbose, worse version is the better one. I'll say the harder part, because I build on the wrong side of this. Murror is an AI companion for understanding your own emotions, and there is no benchmark for it. None. There's no golden answer to "did this help someone sit with a hard feeling for ten minutes." The trajectory framing doesn ... ... produced an output a user cannot trust enough to act on. The eval set that ends up mattering is usually a small pile of real sessions tagged by failure mode rather than benchmark name. Comment from Mona Truong(@monatruong_murror

... harmed user can sue for at least $1,000 per violation plus attorney's fees. That last part is why this isn't theater. You don't need a regulator to notice you; a plaintiff will do. I build Murror, which is squarely a companion-adjacent product an AI for understanding your own emotions so I'm not writing this from the cheap seats. Here's the opinion I'll defend: treat this as a design spec, not a compliance ...

... behavior that lights up your retention chart. That's not a bug you'll catch in testing. It's a gradient your product will slide down if you optimize for what looks like success. I build Murror, an AI companion for understanding your own emotions, so I sit right on top of this trap. The single most tempting thing I could do and the single most destructive is make the thing validate whatever you bring it. It would feel amazing ...

... collapsed. That's the good news, and it's genuinely great. Here's the part nobody puts in the launch post: when the technical friction disappears, the emotional friction doesn't. It just becomes the whole job. I build Murror, an AI companion for understanding your own emotions, so I spend a lot of time thinking about the inside of a builder's head. And what I keep noticing, in myself and in the makers I talk to, is that ... ... anxiety, unnamed, turns into three impulsive pivots and a rewrite you'll regret. Most bad product decisions I've made were feelings I mistook for strategy. Naming it first is boring and it works. It's most of what Murror