Put your mind on your tools
When you forked MemMini you didn’t just copy a framework — you got your mind: your voice, judgment, and knowledge in one portable repo. (MemMini is the framework nobody owns; your fork is yours.) A mind isn’t tied to one app — it runs on surfaces: Claude Code, Codex, a chat app, an agent harness, whatever you use.
Putting your mind on a tool is one step, and it has an easy button.
Just ask
Section titled “Just ask”Tell any agent that can edit your repo:
“Onboard Claude Code as a surface.” (or Cursor, Codex, Aider, a new chat app…)
It runs the agent-bootloader: it figures out what that tool can actually do,
writes down how your mind plugs into it, and tests it in a live session — so the
same who.md / how.md behave correctly there instead of breaking. You say it
once; the agent does every step.
See the map it builds (the glass box)
Section titled “See the map it builds (the glass box)”It never hides the work. For your tool it builds a capability map — your six intentions (acquire, retain, compute, effect, coordinate, defer) matched to that tool’s real buttons, plus a degrade rule for anything it can’t do, so nothing silently breaks:
| You want to… | On this tool, that’s… |
|---|---|
| Effect — publish / deploy | the deploy command — and it pauses for your yes |
| Defer — do it later | a scheduler — or “none here,” written down, never faked |
It shows you each file before it lands and stops for your approval on anything that publishes. Follow along, review, or take the wheel by hand — the same procedure at whatever fidelity you want.
One line to put your mind on a tool, with the map in plain sight. Do it for each tool you use, and the same you shows up everywhere.
- Add a new tool — the step-by-step, when you want to drive
- Write your first rule — change who the version of you is