forest

Inspired by @openclaw - I built Dot - an iOS agent that runs securely on your phone through native api's, app intents, and shortcuts Dot learns new skills by generating shortcuts on the fly - no setup just use the apps you already use Here's Dot learning to talk (sound on🔊)

@tylerangert Bring back spore

can someone pls make a pokemon inspired game where new pokemon evolve independently and the pokedex is infinite

@LuozhuZhang @blakeir eh I hate listening to people speaking - can injest text at least 10x faster

@blakeir voice in, voice out is where it gets interesting

voice in, text out feels like the most natural interface right now

Adding new ios 26 speech analyzer api to Dot w a dedicated speak button - feels so nice already should’ve been in v1

@jacobtechtavern Pain - would be cool to see a chart of keywords added over time Curious how many are concurrency related

Swift has more keywords than all those languages combined 🤯 Go: 25 keywords. Kotlin: 72 keywords. C++: an insane 95 keywords.

@inaplight LMAO yeah after release I’ve come to find there are more than a few sharing the name 😅

@ephemeralapps i have an agent under the same name LOL github.com/plyght/dot

Dot is now managing my emails - crazy how many workflows are unlocked w simple hooks like emails and messages more to come

Added support for Automations (Tasks) Setup once in the iOS shortcuts app and let dot take action on any of your messages or emails “Hey dot when I receive spam email - auto reply with unsubscribe”

@nikitabier @flyosity Request expedited if it’s urgent - I submitted and was approved within 5 hours yesterday

@flyosity It’s been so frustrating. We have 4 solutions to reply spam built—but our app is in purgatory.

iOS developers: How long is App Review taking for everyone these days? It is now taking longer to get our app approved than it is to build the actual features.

@jacobtechtavern Nice really cool - will have to play with it

With Remotion! It lets you define a video timeline via React components I'm still working out exactly how to get them to pop, but my current pipeline is: 1. Convert snippet of my blog post into a script. 2. Feed it into a coding agent that sets it up in re-motion. 3. Nitpick. This is v1. V2 is going to involve my face and chatting to the camera.

Throwback to 2020 where I briefly broke COVID-19 testing in the UK for 59 minutes 💀🦠 I learned the hard way what can go wrong with SSL pinning. (Link in first reply)

@tiborbodecs We love app intents

App Intents... 💪

@Scobleizer I think it just used wine lol - still cool but not like some of the other ports we’ve seen

Someone put a Halo ISO file in a folder, told Codex to make it playable on Mac, left it running, and came back to a working game.

Messages style chat in UIKit built with CollectionView Supports streaming, custom messages, and auto scroll to end Built 100% in codex with gpt 5.3/5.4

Adversarial Interop is gonna be the theme of this year

@edwardsanchez You could just ask 5.4 in codex - make it collection view based - it can make one that feels good and meets all your reqs in a couple hours

I think if someone offered me a ScrollView replacement that handled chat-like (inverse scrolling) behavior in SwiftUI (even UIKit) for $10,000. I'd pay for it gladly. Never have I ever been more frustrated by having to do something seemingly so simple yet at every step of the way there's another gotcha. I think I must have spent over a billion tokens on this problem and nothing.

starting to believe

Will be surprised if someone doesn’t try to recreate this with Codex and GPT 5.4 by end of next week

@twannl 1. Open terminal 2. Enter “codex” 3. Repeat x20

I just hired 20 engineers as an indie developer. Curious how? Read below!

@Clad3815 lol the slow drag to the center of paint after pasting is straight cinema

Nobody seems to know how insane GPT-5.4 is with computer use. I asked GPT-5.4 to draw the OpenAI logo in Microsoft Paint. No computer use API. Just a screenshot and basic tool calls (click, drag, press_key) all coordinate-based. The first drawing was awful. And GPT knew it. It looked at its own result and essentially went "yeah, no." What happened next is what broke my brain: It opened a browser. Went to Bing Images. Searched for the OpenAI logo. Found one. Then (and I cannot stress this enough) it used the Windows area screenshot shortcut (Win+Shift+S) to snip just the logo off the screen. Went back to Paint. Imported it. Centered it. All on its own. No instructions to do any of that. It just improvised a better strategy when the first one failed. My prompt was "Draw the OpenAI logo" with Paint already opened on the computer. Sure, it's "cheating." But honestly? That's exactly what I'd do too. And the fact that it came up with this plan from nothing but a screenshot and a coordinate system is wild.

@faionur @whop Nearly called it x.com/ephemeralapps/…

cluely pivoted to a @whop competitor and you can clip roy lee videos for $$