Memory as files
Your context lives in a folder you can open, edit, back up, or delete.
PersonalOS lets you see, edit, and own the memory your AI agent uses to understand how you think, write, work, and decide.
Nothing becomes memory until you review and approve it.
Your context lives in a folder you can open, edit, back up, or delete.
The agent can suggest what to remember. You decide what becomes durable.
Use the same operating layer with supported agents instead of rebuilding context from zero.
PersonalOS is not a menu of disconnected modes. Its skills use the same reviewed memory, so help can compound across coaching, writing, technical design, job search, decisions, and daily work.
“Use my usual tradeoff lens here.”
“Challenge the pattern I keep repeating.”
“Use my known architecture preferences.”
“Turn my real projects into interview examples.”
“Draft this in a way that sounds like me.”
“Pick up the project context from last time.”
PersonalOS starts simple: a folder, a few profile files, and a first setup conversation. Technical users can add private Git later for diffs and version history.
No. You start by copying two short commands. Technical options are available later.
In your folder. If you sync the folder with a cloud provider, that provider handles the sync.
Claude Code is the first supported runtime. The core is agent-agnostic and open-source.