ultimatixarup/jervis ? reverse-engineered prompt
Reverse engineered prompt
Build me a personal agent for my Mac that I can talk to in plain English and that can actually use the computer for me.
I want to type to it in a terminal first, and have it answer back while showing me what it is doing as it goes. It should be able to read files, look around my Mac, and do useful actions through tools, but anything risky like deleting things should stop and ask me first before continuing. Please make it feel safe, with a clear audit trail of every action it took.
It would be great if I could also talk to it from my phone through Telegram, and later by voice, but typed chat should be the main way in. If you need to look up current docs online for Claude, macOS permissions, or MCP, go ahead. Keep the setup straightforward, with a simple start script and a health check so I can tell when it is ready.
Are you gonna build this?
make sure you review the code using coderabbit