App Claw
Control and manage your personal AI from your phone.
Visit siteAI agents were getting genuinely useful, but you needed a terminal and a free weekend to run one. That left out almost everybody.
When the open-source AI agent wave started, I got obsessed with it.
Then I noticed something. Every single person I knew who was actually using these tools was technical. You needed a terminal, a server, API keys, and enough patience to debug why the thing stopped running at 2 AM.
Everyone else just read about it on Twitter.
App Claw was my attempt to move that whole experience into a phone.
Chat with your agent. See what it's working on. Manage what it remembers. Turn new capabilities on without opening a config file. The complexity still exists underneath, it just shouldn't be your problem.
It didn't go where I wanted it to, so I archived it.
But I still like the problem, and I learned something I wasn't expecting. Taking away the setup isn't enough on its own. People also have to trust the thing before they'll hand it anything that matters.
AI shouldn't need a computer science degree.
What it does
- Chat with your own agent from your phone, with no terminal in the way.
- Watch what it's doing in real time and step in when it goes the wrong way.
- Manage its memory so it holds on to the context that matters to you.
- Add capabilities by switching them on instead of editing config files.
Lesson
Removing the setup is a real product. But people only feel that value after they already trust the thing doing the work.