esa.int
Hubble unexpectedly catches comet breaking uphttps://www.esa.int/Science_Exploration/Space_Science/Hubble_unexpectedly_catches_comet_breaking_upComet K1, whose full name is Comet C/2025 K1 (ATLAS), had just passed its closest approach to the Sun and was heading out of the Solar System. Though it had been intact just days before, K1 fragmented into at least four pieces while the NASA/ESA Hubble Space Telescope was watching. The odds of that happening while Hubble viewed the comet are extraordinarily miniscule.Mar 18, 2026 2:00 PM
bing.com
These are the solutions dermatologists and experts actually recommend for thinning hairhttp://www.bing.com/news/apiclick.aspx?ref=FexRss&aid=&tid=6a8cb9eb19154d1bba6ac4ca17f0d1d6&url=https%3A%2F%2Fwww.msn.com%2Fen-us%2Fhealth%2Fother%2Fthese-are-the-solutions-dermatologists-and-experts-actually-recommend-for-thinning-hair%2Far-AA1YViAM&c=5677297316404691536&mkt=en-usShop TODAY independently determines what we cover and recommend. When you click on or buy through our links, we earn a commission. Learn more. It's an unfortunate fact of life that, for many people, ...Mar 18, 2026 1:28 PM
getcloak.dev
Show HN: Cloak – .env on disk has fakes, your editor shows them (CLI and VSCode)https://getcloak.devI run AI coding agents with full filesystem access daily — Claude Code, Cursor, the works. Last month I realized every one of them has been reading my .env files with real Stripe live keys, database passwords, and AWS credentials. Sent straight to model providers as "context." .gitignore protects git, not disk. Secret managers protect servers, not laptops. Sandboxing the agent kills its usefulness. Nothing stopped a local cat .env from returning real credentials. So I built Cloak. The .env on disk always contains structurally valid fakes — sk_test_ for Stripe, localhost for databases, AWS example keys. Agents read the file, get sandbox values, write perfectly valid code. The VS Code/Cursor extension intercepts file opens and decrypts from an AES-256-GCM vault so you see real values in your editor. On save, it re-encrypts and writes sandbox to disk. cloak run npm start injects real env vars gated behind Touch ID / password — agents can't authenticate. Technical decisions worth discussinMar 18, 2026 1:22 PM
news.ycombinator.com
Roast my first SAAS: Relationship proof builder for visashttps://news.ycombinator.com/item?id=47425423Hi HN, Maco here. I'm the solo founder of VisaAlbum (https://visaalbum.com)I built this after going through the immigration process with my wife twice. Compiling relationship evidence for strict visas (US K-1, UK, Schengen) is incredibly stressful. Couples have to organize hundreds of photos, chat logs, and travel history into consulate-standard A4 PDFs. Generic tools like Word or Canva usually result in massive file sizes, broken formatting, and layout shifts when exporting.VisaAlbum is a micro-SaaS specifically tuned to solve this one workflow.Here are a few technical hurdles I had to solve to make it work in the browser:Browser-Based A4 Rendering: I initially used html2canvas but kept hitting walls with modern CSS. I moved to html-to-image to properly support high-fidelity gradients while respecting exact A4 print dimensions so the final PDF export doesn't shift by a single pixel.Global Media Deduplication: Uploading 300+ images for an immigration album bloats storage fast. I implemMar 18, 2026 1:15 PM
news.ycombinator.com
BirdyChat – A minimalist Telegram-native live chat with voice-to-text featurehttps://news.ycombinator.com/item?id=47423558Hi HN,I built BirdyChat because I wanted a simple live chat for my website, but everything I tried felt like a giant "corporate gargantuan."Also I didn't want any additional app to install to my phone. That's how an idea came to my head - I will create a live chat widget that sends website's messages directly to my Telegram account.How it works: Dashboard in a Mini App: I moved the entire admin UI into a Telegram Mini App. It’s stripped down to just a few essential buttons. No heavy web UIs, no distractions.Voice-to-Text: This is my favorite "productivity hack." I record a voice note in Telegram while on the move, and the visitor gets a clean text message on the site. Fast for me, professional for them.Lightweight Script: Optimized to ensure your site's performance stays high.Zero Registration: Start the bot, paste the snippet, and you’re live.＞＞＞ Try it out (No signup needed): directly in Telegram t.me/birdychat_botOr you can read more in my website: https://birdychat.comI’m curious aMar 18, 2026 9:44 AM
absentkey.com
Show HN: AbsentKey E2E encrypted vault that releases secrets when you can'thttps://absentkey.com/en/My wife doesn't know any of my passwords. If something happened to me tomorrow she'd be locked out of everything, wouldn't know where to find important documents either. I looked at dead man's switch apps and they all wanted me to check in every few days or weeks to prove I'm alive. Miss one check-in and everything gets released, you leak secrets because you were on vacation. Also I really hate unnecessary notifications.So I built something different. In AbsentKey, you share your files/secrets then nothing happens until someone requests access. When they do, I get a notification and can approve or deny. If I don't respond within the timeout I set for that person, they get access automatically. The timer only starts when someone actively asks. No check-ins, no heartbeats.You can set different timeouts per person per secret. My wife gets 3 days. My business partner gets 30. Whatever makes sense.On the crypto side: everything is encrypted on the phone before it goes anywhere. 12-word BIP3Mar 18, 2026 8:15 AM
news.google.com
American Kennel Club Reveals the 5 Most Popular Dog Breeds in the Countryhttps://news.google.com/rss/articles/CBMiakFVX3lxTFBna2lDa3Bhd0NpYk9IZjh4dllZa3dUdTJfdGR1NnA1TWN5dV9CcmktU2ZIemF6RV80TF9WVzFFcjV6Mlc2YUd2Mjg2akxjbmU4S1NzQ2dEbjZ5VndjMXRNX1JMNjRXRkMwQlE?oc=5American Kennel Club Reveals the 5 Most Popular Dog Breeds in the Country TODAY.comMar 18, 2026 7:00 AM
news.google.com
Exclusive: American Kennel Club Reveals Top 5 Dog Breedshttps://news.google.com/rss/articles/CBMinwFBVV95cUxPYVpZV2x6TEU5S2pObS1QY18zRURrbjlzbmpsa0I1cFpvamFLVWhLelc3MHVleGJuTEVrNlFhaTRUdDR4OWthNWwyR2hXQkY5eVNnSFdOakUyVnNhQlZ4RGpBSUZlTzZ3QUNjWF90aVdlYU1SS3lSRHdSeVdLUnhxMm45WjlzV0lzZHZMNzRzM3FPb25jR2Rac1VfY0hGUk0?oc=5Exclusive: American Kennel Club Reveals Top 5 Dog Breeds TODAY.comMar 18, 2026 7:00 AM
bing.com
How Much Does SEO Cost in 2026? Agency Fees, In-House Costs, Tools, and ROI Benchmarkshttp://www.bing.com/news/apiclick.aspx?ref=FexRss&aid=&tid=6a8ccd8efe9c48749ff851785dd9b47f&url=https%3A%2F%2Ffinance.yahoo.com%2Fnews%2Fmuch-does-seo-cost-2026-122000506.html&c=2409170659774477629&mkt=en-usSHERIDAN, WY / ACCESS Newswire / March 18, 2026 / Search engine optimization (SEO) has evolved from a niche marketing tactic into one of the most important long-term growth channels for businesses ...Mar 18, 2026 5:20 AM
news.ycombinator.com
Ask HN: I built a geometric reasoning engine – what should I test it on?https://news.ycombinator.com/item?id=47421835So I fed Euler's identity into my engine yesterday and the results kind of broke my brain.Some context. I'm a web designer from Vilnius. No math degree. Over the past few months I built a geometric synthesis engine (~7,700 lines of Python) that encodes concepts as complex phasor vectors on an E8 lattice and discovers structural relationships through algebraic binding. The LLM doesn't reason — it's just a vocoder that names what the geometry produces. All structure is deterministic. Same input, same output, every time.The mode I want to show is Reverse Genesis. You give it a goal, it decomposes the structural prerequisites — what has to be true before this outcome is possible.I pointed it at e^(iπ) + 1 = 0. It produced 27 prerequisites.Things like "Spinor Sign Flip — a half-rotation in a hidden spin space to align phase factors for cancellation." That's... mathematically correct. e^(iπ) is a π rotation, which in spinor algebra is literally a sign flip."Root Collapse Path — collapsing thMar 18, 2026 5:19 AM
bing.com
Volaris Expands Workforce Management Presence with Acquisition of Comprosehttp://www.bing.com/news/apiclick.aspx?ref=FexRss&aid=&tid=6a88f5376e124dcaafe0586ec0344c6f&url=https%3A%2F%2Ffinance.yahoo.com%2Fnews%2Fvolaris-expands-workforce-management-presence-120000287.html&c=6593950581796646332&mkt=en-usTORONTO, March 18, 2026 (GLOBE NEWSWIRE)-- Volaris Group today announced the acquisition of Comprose, Inc. ("Comprose"), a provider of policy and procedure management software for enterprise customers ...Mar 18, 2026 5:00 AM
bing.com
Volaris Expands Workforce Management Presence with Acquisition of Comprosehttp://www.bing.com/news/apiclick.aspx?ref=FexRss&aid=&tid=6a8c3abd884d4e979732569857f16fd1&url=https%3A%2F%2Ffinance.yahoo.com%2Fnews%2Fvolaris-expands-workforce-management-presence-120000287.html&c=6593950581796646332&mkt=en-usTORONTO, March 18, 2026 (GLOBE NEWSWIRE)-- Volaris Group today announced the acquisition of Comprose, Inc. ("Comprose"), a provider of policy and procedure management software for enterprise customers ...Mar 18, 2026 5:00 AM
news.ycombinator.com
TrustAgentAI – Cryptographic receipts for MCP tool calls (non-repudiation layer)https://news.ycombinator.com/item?id=47421489MCP solves connectivity between agents and tools. It doesn't solve accountability.When an AI agent calls execute_wire_transfer today, there's no cryptographic proof it happened, no proof who authorized it, and no tamper-evident record of the outcome. Editable logs and unverified timestamps don't cut it when real money or infrastructure is on the line.I've been building TrustAgentAI — an open-source accountability layer that wraps MCP tool calls in a 3-phase signed receipt protocol:Intent Envelope → Agent A signs: "I intend to call X with these exact args" Acceptance Receipt → Agent B signs: "I validated and accepted this intent" Execution Envelope → Agent B signs: "I executed it, here is the outcome hash"Each receipt is signed with Ed25519 over a JCS (RFC 8785) canonical hash. All three are chained in a DAG ledger — causality is cryptographically enforced, not just logged. The ledger Merkle-batches entries and anchors roots externally (L2 blockchain), making retroactive tampering detecMar 18, 2026 4:04 AM
nasa.gov
NASA’s X-59 Prepares for Second Flighthttps://www.nasa.gov/centers-and-facilities/armstrong/nasas-x-59-prepares-for-second-flight/NASA’s X-59 experimental aircraft is preparing for its second flight, a step that will set the pace for more flight testing in 2026. Over the coming months, NASA will take the quiet supersonic jet faster and higher, while validating safety and performance, a process known as envelope expansion. NASA test pilot Jim “Clue” Less will be at the X-59’s […]Mar 17, 2026 8:55 PM
spacenews.com
Office of Space Commerce weighing options for TraCSS user feeshttps://spacenews.com/office-of-space-commerce-weighing-options-for-tracss-user-fees/The Trump administration has not yet decided whether to charge for space safety data despite a change in space policy enabling the government to do so. The post Office of Space Commerce weighing options for TraCSS user fees appeared first on SpaceNews.Mar 17, 2026 8:48 PM
news.ycombinator.com
Launch HN: Kita (YC W26) – Automate credit review in emerging marketshttps://news.ycombinator.com/item?id=47417335Hey HN! We’re Carmel and Rhea, the founders of Kita (https://www.usekita.com/). We automate credit review for lenders in emerging markets using VLMs.In many emerging markets, like the Philippines and Mexico, credit infrastructure is weak. Open finance is still nascent, and credit bureaus are unreliable. So to apply for a loan, lenders rely on borrowers submitting documentation to understand their ability to repay. A borrower can submit financial documents, such as bank statements and payslips, in any format, from pdfs, images of physical documents and screenshots. On top of that, financial documents in these markets are highly unstandardized, with no consistent templates lenders can rely on.Existing OCR and document AI tools break on these highly variant, messy real-world documents. Generic tools are not built for lending workflows like verification, fraud detection, and risk extraction. As a result, credit teams fall back on manual review, making underwriting slower, more expensive, aMar 17, 2026 7:46 PM
brrrrfect.com
Show HN: I built a rental property underwriting app and used it to buy a househttps://brrrrfect.com/Over the last few months, I built Brrrrfect.com as a way to teach myself how to underwrite properties. I've always had a relative disdain for spreadsheets and don't want to give Microsoft money for its office tools, so I decided to build this.I think it would have taken way longer for me to get the hang of underwriting real estate if I hadn't spent time building this and really learning about what the calculations mean.I ended up using this tool to buy my first Buy, Renovate, Rent, Refinance, Repeat (BRRRR) deal and its working out really well! I'm definitely going to keep using it for myself.If it helped me, I figured it might help someone else, so I made it live! Right now it only supports basic residential underwriting and a more advanced BRRRR calculator, but I want to add commercial and farm land underwriting tools in the future as well.I’d love feedback from people who invest in real estate, underwrite deals regularly, or just have opinions on what could make this tool more usefuMar 17, 2026 6:01 PM
github.com
Show HN: One, cross domain auto-researching knowledge graph Claude orchestratorhttps://github.com/bcd532/onehi, so straight to the point. i had claude code $20 for a while, and before upgrading i was always thinking about a way to make an "infinite context system", i also work... A LOT. 22hrs a day or so?so i worked around, did a lot of trying with mcp, plugins, and i stuck with a system i call "one".hdc vector embeddings (4096 dimensions, trigram + word encoding) stored in SQLite and recalled by cosine similarity on context shifts.entity extraction builds a knowledge graph across sessions. rules get learned from repeated preferences. thats the corethe part that scared me was the autonomous research loop. there's a mode where claude researches a topic, then a dialectic engine challenges every finding. thesis/antithesis/synthesis.a contradiction minor looks for conflicts, and a synthesis engine searches for patterns across domains. weak findings get pruned and it can iterate indefinitely.it was running on my 15m kalshi trading algorithm (which also happens to use hdc + tsetlin machines haha!)Mar 17, 2026 5:23 PM
credibilityprofile.com
Show HN: CredibilityProfile – bringing reference checks earlier into hiringhttps://www.credibilityprofile.com/Hi HN,I’ve been a startup operator and advisor for 10+ years, and one recurring frustration has been hiring, which got even worse with the introduction of some new AI tools: CVs are too polished, too many applications to screen, interview capacity is limited, real signal (how someone actually works) comes late via references — if at all.So we built a small tool: https://credibilityprofile.comIt lets candidates collect structured feedback from past managers/peers (with consent to be contacted), and share it as a reusable link in their CV.We ran an initial pilot in our hiring processes, and seen some promising results: Candidates were willing to complete and then re-use it in their CVs for their other applications.Right now, we’re testing it with other entrepreneurs/hiring managers/recruiters — with a very simple setup, at no cost (just adding a note to job posts).Open questions we’re still thinking about:- Is this meaningfully better than LinkedIn recommendations / references later in pMar 17, 2026 5:06 PM
punkcode.rocks
Show HN: PUNK – Remote control for local Claude Code that just workshttps://punkcode.rocksMy laptop has accumulated a lot of state.It's logged into everything. GitHub CLI, Slack, email. Repos are already cloned. Every day it gets richer just by being my laptop.I can't really move this into the cloud without losing something. So I stopped trying. I just wanted to reach it from my phone.At first it was just for me. Before TestFlight I was installing it by plugging people's phones directly into my laptop. It spread around South Park Commons and to engineers at Bay Area companies who were already using Claude Code. Now about 50 people use it.A few ways I actually use it:Sometimes I wake up at 2AM with something clicking — a bug, a feature idea, a thought I don't want to lose. I whisper to PUNK and go back to sleep.In the morning I close the laptop, throw it in my bag, and walk to the office. About 10-15 minutes. The agents are still running the whole way. I'm on my phone checking progress, redirecting things, starting something new. By the time I walk in, they've been working tMar 17, 2026 3:47 PM
