Agents At Work

Monitor and control AI agents from your phone

19 followers

Monitor and control AI agents from your phone

19 followers

Monitor and control Claude Code, Codex, Gemini CLI, and Grok AI agents from your iPhone or Android. Push notifications, one-tap approvals, live conversation feed, from anywhere.

Payment Required

Launch tags:Android•Mac•Developer Tools

Launch Team / Built With

Marked 3Writing to LLMs, Marked is the ultimate Markdown preview.

Promoted

UpWrite AI: Proofreading Keyboard

Maker

📌

Hey hunters! 👋 I'm Sergey, the maker of Agents At Work. The problem: I kick off Claude Code on long-running tasks, then walk away from my Mac. Every time, the same thing happened. I'd come back an hour later to find the agent stuck on a permission prompt it asked ten minutes in, just sitting there waiting for a yes/no I never saw. Or it had crashed silently. The "let it run autonomously" promise of these CLIs falls apart the moment you're not at the keyboard to babysit them. I actually started building this for myself before Claude shipped its own remote control and mobile features. I figured Anthropic would eventually build something like it, and they did. But once I had my own bridge running, it was already doing more: full conversation context, detailed permission prompts, and it wasn't locked to one agent. I'm constantly switching between Claude Code, Codex, Gemini, and Grok depending on the task, and I didn't want a separate babysitting workflow for each one. That's when I decided to generalize it and make it public instead of keeping it to myself. A few things I looked at before building this: - Claude.ai mobile lets you dispatch tasks to Claude while your Mac is awake, which is great for simple jobs. But it only controls Claude, gives limited real-time visibility, has simpler permission prompts, and no conversation feed. No Codex, Grok, or Gemini. - Generic push tools (Pushover, ntfy, etc.) just send a notification. No context, no way to reply or approve from the alert itself. - OpenClaw does remote control over Telegram for quick commands, but setup is involved and there's no dedicated UI showing full session context. So I built the bridge I wanted. Hooks on the agent side write events straight to Firestore, a small daemon keeps a heartbeat and listens for commands, and the phone app shows a real conversation feed instead of a beep. When the agent hits a tool call that needs approval, you get a push notification with the actual question and tappable options. Answer from your phone, and the daemon sends the keystroke back into the terminal session so the agent continues. You can also drop a note mid-session, like "skip the tests for now," without opening a laptop. End-to-end encryption (AES-256-GCM) went in once I realized I was piping real prompts and file paths through a third-party backend, and I wasn't comfortable with that myself. If I wouldn't use it that way, I wasn't going to ship it to anyone else. Right now it supports Claude Code, Codex CLI, Gemini CLI, and Grok CLI, with native iOS, Android, and macOS apps. Auth is QR-only on mobile: scan and go, no passwords. How do you handle unattended agent sessions today? Happy to dig into the architecture, the encryption approach, or why I went with four agents instead of just one.

Report

12d ago

UpWrite AI: Proofreading Keyboard

Maker

Hey all! Shipping a big update today.

You can now run local AI models through Ollama and control them from your phone the same way you control Claude, Codex, or Gemini. Pick a project folder, pick a model (qwen2.5, qwen3, llama3, mistral, or whatever you have pulled), and send prompts from your phone. Responses stream back through the same conversation feed. No API key. No subscription per query. The model runs on your Mac; your phone is the remote.

This matters for a few reasons:

Some tasks work better with a fast, local model than a frontier API. Writing, brainstorming, and anything you would normally batch overnight are good examples. You do not always need the most powerful model available; sometimes you just need something that responds instantly, privately, and without counting tokens. Ollama makes that possible, and Agents At Work makes it controllable from anywhere.

It also opens up something genuinely new: you can now have Claude working on a coding task, Gemini on documentation, and Ollama on a writing project, all visible from the same app on your phone at once. The conversation feed, push notifications, and remote prompting work identically regardless of which agent is running.

One thing worth saying: a lot of people assumed this tool was only for developers because the early adopters happen to be developers. But local Ollama support makes that even less true. Anyone who works on long-running tasks and wants to step away from their desk without losing the thread can use this now. You do not need to write code for it to be useful.

macOS companion, iOS, and Android updates are all live.

Report

3d ago

Sounds exactly like what I wanted. I know some of the agents already support this, but not all to the same feature set and I bounce around between my agents a lot. Excited to see how this turns out!

Report

9d ago

UpWrite AI: Proofreading Keyboard

Maker

@evilmikedg Thanks Mike! That mismatch is exactly why this ended up supporting four agents instead of one, and more to come!

Curious which agents you bounce between most, and if you try it, I'd love to hear what breaks.

Report

9d ago

@serge_neskoromny Gemini for sure.

If you ever got Amazon kiro, id try that for sure too.

At some point would probably try the others since I do tend to bounce around every couple of months. But those 2 are what my current subs are

Report

7d ago

Our phone is the sun of our personal device solar system, having the remote control there makes so much sense, @serge_neskoromny

Report

9d ago

Reviews

Hey all! Shipping a big update today. You can now run local AI models through Ollama and control them from your phone the same way you control Claude, Codex, or Gemini. Pick a project folder, pick a model (qwen2.5, qwen3, llama3, mistral, or whatever you have pulled), and send prompts from your phone. Responses stream back through the same conversation feed. No API key. No subscription per query. The model runs on your Mac; your phone is the remote. This matters for a few reasons: Some tasks work better with a fast, local model than a frontier API. Writing, brainstorming, and anything you would normally batch overnight are good examples. You do not always need the most powerful model available; sometimes you just need something that responds instantly, privately, and without counting tokens. Ollama makes that possible, and Agents At Work makes it controllable from anywhere. It also opens up something genuinely new: you can now have Claude working on a coding task, Gemini on documentation, and Ollama on a writing project, all visible from the same app on your phone at once. The conversation feed, push notifications, and remote prompting work identically regardless of which agent is running. One thing worth saying: a lot of people assumed this tool was only for developers because the early adopters happen to be developers. But local Ollama support makes that even less true. Anyone who works on long-running tasks and wants to step away from their desk without losing the thread can use this now. You do not need to write code for it to be useful. macOS companion, iOS, and Android updates are all live.