anil-matcha/open-dots ? reverse-engineered prompt
Reverse engineered prompt
Build me a self hosted personal AI agent workspace that feels like a local first ChatGPT style app, but with more control.
I want chat with saved conversations, markdown, image attachments, and a few assistant personas that can each have their own instructions and model settings. Add a settings area where I can plug in my own model provider, pick a default model, and save the API key locally. The app should support tool use with a clear approval step before anything risky happens, like reading or writing files, searching the web, or doing computer actions. Please include a simple audit trail so I can see what the agent tried to do and what got approved.
It would also be great if the app can connect to services like Composio, and optionally run browser based computer tasks with a local runtime or a remote computer service. Keep the data local, use SQLite, and make credentials encrypted. If you need to check current docs for any libraries or browser automation pieces, go ahead and look them up online.
Are you gonna build this?
make sure you review the code using coderabbit