alva
@filipealva
10+ years developing for iOS, making #AI apps for people outside the bubble. Building http://writesuit.com
animations aren't my strength, but working on a pretty cool one now i had to spin off a new project just so i can make it right without all the noise and then will migrate it back in
Oh yeah… and basic security is the moat The bar is low on that one at the moment
The only trade secret is user acquisition now What keywords you’re targeting What marketing you’re doing This has been the biggest shift I’ve seen in my entire career… Technology can now be copied in an instant now. But without users it’s just code
before and after of @writesuit home page designs fully implemented by @brunosquadros using @cursor_ai without previous coding experience can't wait to see the whole product with the new design highlight for the new button styles 🔥
the tea app hack is not a loss for ai coding its a reminder that you can use all the tools available to ship fast but that doesn’t mean you shouldn’t be diligent
go check namesnag out, i got writetopersuade .com there that drives 800 visits weekly to my ai writing product
I didn't fully appreciate how many domains become available each day. I can't keep track! Here are some things I've added to try and help: - ✨New✨ marker on recently added domains - Sorting works now (was only sorting by score) - Added filter/sort by recency (X days, "recent")
i love this change
traditional software development has flipped. BEFORE: designers visualized products as mockups, engineers had to clean up edge cases & make it all work. NOW: engineers vibe code features so fast that designers are cleaning up UX to make them work with the product.
While everyone is freaking out over sub agents, I’m here making collaborating agents
I've long said that o3 is the best coding model - but if you're using an agent harness - Claude is just better at navigating your codebase. Enter the @RepoPrompt pair programmer mode - it's the best of both words, as Claude coordinates with o3 to plan and apply edits for you!
hey @nikitabier a gist feature would be a great addition to x the everything app
Enjoy! public struct GlassCardModifier: ViewModifier { public let radius: CGFloat public init(radius: CGFloat = 16) { self.radius = radius } public func body(content: Content) -> some View { content .background { if…
true i’m a big fan of hiring based on what people build, seeing the code, chatting about it and then a trial period would save us a lot of time and bullshit if it became the industry standard
Idk what’s worse, interviewing for roles or being the interviewer. The whole process is painful either way.
accurate
My take: July 2025 LLM coding meta - o3 for research/planning - o3 for code review - o3 for debugging - sonnet 4 for agent tool calling driver - o3 for first pass code implementation - sonnet 4 for iterative changes to existing code - Gemini 2.5 Pro for long context edge cases
my @cursor_ai bugbot trial summary: 100 prs reviewed 86 bugs found 82% of them fixed
One of the advantages of async/remote is that almost everything is written. With AI, the answer to any forgotten question is just 30s away:
Usage analytics is live -- find out what your users are really doing, and let us know what else you want to see @amianthus_ you cooked🫵🧑🍳