Personal context file
Why it exists
Every new AI chat starts cold. The model does not know your role, your stack, your voice, or what you already told another assistant. Built-in memory helps a little, but it is trapped inside one app. A personal context file lives outside any single tool: one page, every agent.
With vs without personal context
Illustrative · not a formal benchmark
What goes in it
- Identity: who you are, in a few lines an AI should never get wrong.
- Goals: what you are working toward right now.
- Work: your role, stack, projects, and how you operate.
- Preferences: how you want AI to talk, format, and assume.
- Routines: the shape of your week that suggestions should respect.
- Optional: Beliefs, Constraints, People, Health, Context, when they earn a place.
Specific over complete. If a line would not change an answer, leave it out.
Context file vs chatbot memory
| Topic | Chatbot memory | Personal context file |
|---|---|---|
| Where it lives | Inside one app | One file you own |
| Portable across tools | No | Yes |
| You can read and edit it | Partly | Yes, plain Markdown |
| Export or delete | Limited | Anytime |
| Kept current by | Opaque heuristics | Agent proposals you approve |
How it stays current
Connected agents read the file before they answer and propose small updates as they learn something durable. You approve what stays. Session chatter stays out.
Common questions
One file that describes who you are and how you want AI to respond, which every connected tool reads before it answers.
ChatGPT memory lives inside ChatGPT. A personal context file is portable Markdown you own, read by every agent you connect.
Durable facts that change how AI should respond: identity, goals, work, preferences, constraints. Leave out session chatter.

