HN user
Nice. Obviously this is a huge issue many are trying to solve. Providing agents with a keychain is one thing, but how does this solve the issue of using these keys in projects? Does the agent decrypt it and set it in a .env? Is the CLI required to be used in the build process?
good point, done!
Building upfile.sh because AI agents need a better, cleaner way to upload screenshots and files, in order to get a URL back instantly.
Supports public (permanent CDN), expiring (TTL-based), and private (auth-gated) uploads. --json flag for scripting/agents.
Built on Cloudflare Workers + R2. Currently waitlist, happy to give HN folks early access.
If you can submit a screenshot or crash report via test flight that would help. There’s also a Debug option in settings. It lets you email those logs to me, which would be helpful
You can add as many as you’d like.
If you swipe on a station, it lets you add as a favorite
Right now I’m doing it simply by inserting history into a DB and calculating it just by the amount of times in the last 60 days. I plan to do more complex statistics later on, I think there’s more to do there.
I was pleasantly surprised with how both Codex and Claude helped me with this. I’ve been meaning to learn ios dev for the longest time and I’m really glad I finally managed to give it a try while having enough help doing it! I’m a React developer for the most part.
I’ve been building an iOS app called ViaNJ, an unofficial real-time NJ Transit companion. It’s my first iOS project and I’ve been enjoying playing around with Xcode (and various AIs to help out).
It shows live departures and mobile ticket integration with Apple Wallet passes, track predictions, with plans for alerts, GPS-based train map, Live Activities… and more.
I’d love help from the community testing it, finding bugs, and suggesting improvements. It’s been a fun side project, and I’m looking for feedback on what features matter most to daily commuters.
TestFlight: https://testflight.apple.com/join/da4zcqfm Feedback & roadmap: https://feedback.vianj.app
I really like this! At first glance it reminds me of what Apple did with Foundation Models and Guided Generation. Definitely aiming to try it out.
Maybe they changed their pricing but my team paid $25 for 25,000 validations as a one time thing. We’re not in need of 10,000 a month so I wouldn’t be able to try it out soon. But yeah I guess I’m just still not sure how it’s that different from Emailable. Not that there’s any issue with competition of course!
Sounds really useful. Still, I believe Emailable does that, including the scores. Have you tried it?
Neat. What are you using on the backend to verify? How is it different from other platforms such as Emailable?
Reminds me of that old Apple Watch feature that’s been lost to time https://www.nbcnews.com/tech/gadgets/heart-heart-apple-watch...
Since I could not find any examples of this online, and had to figure it out, I wanted to share it online.
This React TypeScript project demonstrates a Braintree OAuth integration flow between two merchants. It showcases how Merchant A can obtain authorization from Merchant B to perform actions on their behalf using Braintree's OAuth and transaction APIs.
In this example, Merchant A provides this application to Merchant B, allowing them to:
Grant OAuth permissions to Merchant A
Allow access to customer payment methods
Enable facilitated transactions
The integration uses two main Braintree APIs:
Access Token API - For OAuth authentication
Transaction API - For creating Shared Vault Transactions
FEATURES
OAuth connection flow
Token management and storage
Shared vault transaction processing
Real-time transaction status updates
Error handling and user feedback
Hope this is helpful to someone!
Confused, this looks more like a tool that pulls existing HTML from other websites. Is the idea to create your component library by cloning existing components that are out there and modifying them? And then they are hosted in this cloud library of sorts?
This actually has a lot of potential I think. It reminds me of https://bit.dev (obviously without most of the main features). But the core is there - the idea that components should be organized in a way that makes them easily reusable.
If you were to add a preview functionality of each component, this would already look A LOT like the Bit platform.
Personally, I've been trying to figure out a way to create a component library where my components are easily shareable and importable, but haven't yet come to a real solution. Bit is a good one but I am not yet ready to pay per component after my 20 free components are used up.
I think this would be actually extremely useful to me personally if you added these two things:
- component live preview (perhaps with mock props/data?)
- a way to store the components in the cloud so that they can be important into any project, where the source of truth is in this cloud platform
In the future, you could add things like:
- component-level testing
- AI features
- version control and PR reviews
- CI integrations / other integrations
Amazing if YC backs this one too
Ever hear about Levels by Derrick Reimer? Seems like this is trying to accomplish the same thing. He eventually shut it down though. Looks like you got YC funding though so that’s huge!
Damn, this is SUCH a nice UI. Well done! Can't wait for Microsoft and iCloud :)
D'oh, yes, I do mean Notepad++. I haven't used Windows in a while
But good point, I totally forgot about BBEdit
Kinda crazy that it took this long to get a Note++ equivalent (or similar) for macOS! Until now I've been using Sublime for all my quick dev-related scratch notes. This should make things way way better!!
I agree re the app icon - it would look great and still unique by simply removing the name.
Looks cool! Logo seems… Linear-ular… lol
LOL ikr
Any chance you'd open source this? I'm sure most of us won't want to literally share our AirTag with a random email address (even though we would give it a random name).
Either way, seems cool, would love to try it out as an open source project.
No I think you are right. I guess in this regard it really is not so different after all. Thank you for your responses, I will definitely think about migrating over from the iframe I am using.
Good point, I guess that’s true. So regarding your users being able to see your raw JS? What helped that decision be made? I assume that since it’s not your actual React code that makes it a bit better? Sorta?
This is actually really interesting that you switched over to using Web Components instead. Makes total sense for performance and other reasons, of course.
But does this basically mean that you now provide the entirety of the code to users? This includes HTML and JS, since it now lives entirely on the host site, easily accessible to the developers of the host site. Is this something that was discussed and was eventually agreed upon? I would be curious to learn why it was ultimately decided that you are OK with users having the raw HTML and JS code that powers your product.
Furthermore, it also makes it more difficult for users to receive UI updates. If there is an important bug fix or enhancement, users would need to upgrade to the latest HTML/JS "package" that controls the web component, correct?
Of course with an iframe, any updates are seen immediately on the user side.
Curious to hear your thoughts!
Technically this already exists. There's Amazon Subscribe & Save (and similar) for this reason, pretty much. It may not be a subscription like you are describing, but it accomplishes the goal: set it and forget it for consistent grocery (or other) purchases, and get a nice discount.
I wonder if a legit subscription service, for perhaps a select group of items that a user consistently purchases, would work. Could be complicated when things like eggs fluctuate so much in price.
We just launched our end-to-end fully online insurance get-a-quote experience! Give it a try and don’t hesitate to reach out with any questions. https://www.coverdash.com