@twodayslate@twostraws mines been sitting for like 5 days which is unusual for an update. Curious if Apple segments based on app category? I'm developer tools. which category was yours?
Definitely feel like the AI companies are just trying to increase token count (๐ฐ๐ฐ๐ฐ) with the whole Fable/โdonโt pay attention to your agentโ thing. When Iโm working I PREFER to pay attention. I canโt manage 10 different things at once (and have any idea whatโs going on)
Sneak peek of an unreleased Remote Codetrol feature.
Multi-agent build on a headless Pi โ live preview on my MacBook. No tunnels, no config.
Also works on mobile.
Try that with the Claude app.
https://t.co/iQOZ1JowAD
Most apps treat the iPad like a big-screen iPhone.
Remote Codetrol just released an iPad-optimized workspace for managing AI coding agents from anywhere.
Phone, iPad, Mac. Pick your screen!
https://t.co/iQOZ1JowAD
When Claude sends you a wall of text, how do you pinpoint exactly what to change? With Claude Desktop, you get a text field. Good luck.
Remote Codetrol lets you leave feedback inline โ on responses or any markdown file. More precision, better output.
https://t.co/iQOZ1JowAD
Claude Desktopโs autosend lets you queue ONE thing. Got a laundry list? It batches your next two tasks into one message. Thatโs not a queue. Thatโs a pile.
Remote Codetrol sends each task atomically after the previous finishes. No batching. No surprises.
https://t.co/iQOZ1JowAD
Every tool call your agent makes is tracked. Strict permissions, full YOLO, or anywhere in between โ and you can hot-reload mid-session without killing anything.
Same muscle memory across every supported provider. Trust on your terms.
https://t.co/iQOZ1JowAD
Claudeโs โRemote Controlโ only works one device at a time, and you canโt start sessions from mobile. Other tools route through Telegram or WhatsApp.
Remote Codetrol gives you full access from any device. Start sessions from your phone. End-to-end encrypted, no middlemen.
Why marry one provider? Remote Codetrol works with Claude, Codex, Copilot, Gemini, Ollama, GLM, MiniMax, OpenClaw โ Kimi and Perplexity coming soon.
One vendor goes down? Add another agent to your session and keep the work rolling.
https://t.co/iQOZ1JowAD
The workflows I create are fairly vanilla and lean very much "human-in-the-loop". But the cool thing is that you can give your agent whatever skills you want it to have and then build out extremely powerful workflows that can automate...just about anything.
I still like to press the buttons and dictate what the agents do. I keep them on a pretty tight leash. I don't do a ton of MCPs or pull in a ton of agent skills. I pull in the minimum tools I need for the task and prefer chaining "testable" scripts with minimal agent discretion.
AI agents are pretty decent at running tasks, but real work comes with real decisions, and I'm a little bit on the paranoid side when it comes to letting the agents do too much.
Last week made a pretty big update to @remotecodetrol
You can now build full workflows--with branching, input, and a live tracker. In true RCtrl fashion, you can run (and build) them from your phone. Also included a "built-in" agent skill so your agent can build them for you.
This is why you have to be careful with coding agents:
(This is iOS specific)
Working on a feature for Remote Codetrol I asked it to preserve the state of a view when I navigated forward or changed tabs.