Launched this week
Port22
Claude Code, Codex & more on your phone
320 followers
Claude Code, Codex & more on your phone
320 followers
I'd start a long agent run, walk away, and come back to find it had spent 20 minutes waiting on me to approve one file edit. Port22 puts every coding agent running on your Mac onto your phone. See which are working and which are stuck. When one needs permission your phone buzzes and you tap the actual option it offered, not a guessed keystroke. It attaches to what you already run. No wrapper, no config, no new terminal. Free for one Mac and two sessions, every feature on.
@harsha_chaganti Congrats on the launch. Coming back to the Mac and finding it sitting there waiting for 20 minutes is exactly the thing.
I've built a little dashboard for my own sessions, so I'm curious about one part: approving from the phone makes you quicker and less careful than at your desk. Have you seen that in your own use, or does the approval card give you enough to decide properly?
@martin_herran Hey, thank you. And yeah, i do move faster on the phone. Thats real,
i've caught myself approving things i would have read twice at my desk.
Which is why the card has to carry enough on its own. Showing the exact question the agent asks with its real options as buttons, what it still misses and something i have been working on is the repo and the command sitting right next to it which to an extent shows up but still differs for different types of terminals being used
Curious to know what your dashboard looks like or reads from.
@harsha_chaganti Quick question from someone who lives in long-running agent sessions:
When your phone buzzes for approval, how much context do you actually see before tapping?
For example, if Claude wants to edit 3 files or run a risky command, do you get a preview of the exact changes, the ability to jump into the live terminal view before approving or is it mostly “approve/deny” with minimal context to keep it fast?
@swati_paliwal Its not approve/deny. You get the exact question the agent asked with its
real options as buttons, so a six option prompt shows six buttons, not my
guess at yes/no.
Above it you get the conversation, so you can read what it was doing right
before it asked.
What you dont get yet is that rendered as a proper diff on the approval card
itself, or a jump into the live terminal from that screen. Both are on the
list and i am working on showing the diff part with the green and red highlights as of now.
@harsha_chaganti Congrats on the launch! 🚀
I use Claude Code every day and I've definitely come back to my desk only to realize it had been waiting for an approval the whole time 😅
Really like the idea of keeping long-running agent sessions moving without changing the existing workflow.
@ir3ne Thank you. Thats exactly the moment i built it for, walking back and
realising it had been sitting there the whole time.
And yeah, not changing the workflow was the point. It attaches to whatever
youre already running instead of asking you to run it differently.
Reading the options off the live session instead of guessing a keystroke is the part that makes this usable. What I'd still get wrong on a phone is approving the right button for the wrong reason, because I can see "edit src/auth.ts" but not what's actually in the diff. If the approval card carries enough of the change to judge it, I'll clear a queue of these from a coffee shop. If it doesn't, I walk back to the Mac anyway and the buzz just moved the interruption earlier.
@asadmalik901 Really glad you asked that question, and i am working completely on that. As of now you can see the code that changed when you click on the tool call/Edit.
@harsha_chaganti that's the bit that decides whether I keep it installed. one tap to expand is fine, two and I'll just approve blind, which is the exact failure mode I'm trying to avoid. if the diff is under ~20 lines, show it on the card by default and let me scroll. tapping to find out what I'm approving is a tap too many.
@asadmalik901 Absolutely, wont disappoint anyone through the journey. Bare with me and keep supporting as soon as the app is out during the Beta and the first Stable launch will be worth it with all the suggestions and feedback i have received or will receive!
@harsha_chaganti put me on the beta list. I'll run it against a real four hour session and tell you exactly where the approval card wasn't enough, which is worth more to you than me saying nice things here.
The approval card answers the trust question well. What I'd want to know is the network side, is the phone talking to the Mac directly over the same wifi, or through a relay so it still works once you've actually left the house and you're on cellular? That's usually where local dev tools like this quietly stop working right when you need them most.
@raffay_sajjad Hey, to answer your question it is both. The phone talks to the Mac directly over the same wifi and when you are off-LAN or off the same wifi it falls over to a relay so the same experience over cellular as well!
@harsha_chaganti Good, that covers it. Direct when it can, relay when it can't, is the right default and matches how people will actually be using their phone for this.
Tapping the actual option it offered instead of guessing a keystroke is the detail that sells this. Most remote-approval tools just fire an enter key and hope.
@irahimiam The enter-key approach is worse than unreliable, it fails invisibly. Fire enter at a six option prompt and it takes option one, the session keeps moving, and nothing anywhere says a choice was made for you. You only find out an hour later when the run went somewhere you never asked for. That is the real argument for reading the options off the live session. Not that guessing fails, but that guessing succeeds loudly and wrongly.
the way it surfaces the actual approval prompt instead of making you guess at keystrokes is a really thoughtful touch. feels like you built it around the exact moment of frustration rather than building a dashboard first.
@violet_rockefeller Yes, i did use the native apps but i am a long session multi provider user so wanted something and actually went with it, seems like a problem most have i guess
FetchSandbox
The actual options from the live session feels like the killer detail here. Not just “agent on your phone,” but the part where you can safely unblock it without guessing what’s happening back on the Mac. Also love the Port 22 name. Super clean concept, congrats on the launch.
This is exactly the kind of thing I wish I had last week when an agent burned an hour waiting on a yes/no I never saw. One thing that would make it even better: let me set a default action per agent or per project for common prompts like file edits under a certain size, so my phone only buzzes when something actually needs my judgment instead of every little change.
@calvin_sally That is a genuine feature actually, almost like something between auto mode and manual one
Makes sense, thanks for that!