Query me

How it works

Where your memory actually lives.

This is the working-structure model packaged in Query me Starter—not automation performed by the free bootstrap. It centers on a folder you own, an index, and explicit rules that a configured AI agent follows.

Product boundary: the free bootstrap intentionally contains only a minimal reference. Query me Starter packages the working protocol and AI-guided self-setup. QMDrive and QMForge are separate products in development. This diagram explains agent behavior under the rules; the files are not an always-running service.

A folder on your Drive

Plain documents in your own account. Not our server. We do not run one.

You own this

Query me

A structure and set of rules for keeping a current version and opening the note that matters to a question.

  • Your approved tool
  • Exact setup matters
  • Verify before trusting

↻  useful new context can go back into the same folder after review

  1. 01AskYour question starts without your whole archive attached.
  2. 02FindThe index opens the one relevant note.
  3. 03SaveUseful new context goes back to the note you own.
The loop is the point: reviewed context can return to a folder you control. A tool-specific workflow must be tested before it is relied on. See the evidence page for the current validation limits.

What it keeps

A shape you can actually picture.

This is roughly what a setup folder looks like after a few weeks of ordinary use. Every line is a document or folder you can open.

In your folder
What it is for
When it is used
index.md
The table of contents. It lets an assistant open the note that matters instead of reading the whole folder.
Every question
about-you.md
The handful of things that stay true for years: your role, your company, and constraints that rarely move.
Current
preferences.md
How you want work done: tone, format, level of detail, and things you do not want to explain twice.
Current
projects/q4-launch.md
One document per ongoing thing: decisions, constraints, deadlines, and what is still undecided.
When relevant
people/priya.md
Who somebody is to you and what you have agreed with them: context that makes a draft email sound right.
When relevant
archive/
What used to be true. It stays only while it still explains a decision, then can be retired.
Older notes
skills/
Reusable ways of working, such as how you run a review or draft a proposal. They open only when a task needs one.
When needed
mailbox/
Short notes your assistants leave each other so a change made by one does not surprise the others.
For handoffs

Underneath

The parts that are easy to miss.

Storing text is the easy half. These are the habits that keep a memory useful after the first month.

It knows what is still true

A memory that only ever adds eventually contradicts itself. Query me separates what is true now from what used to be true, and keeps the old version only while it still explains a decision.

It cleans itself up on purpose

Detail that has stopped earning its place can be shortened or retired on a schedule, with a small record of what changed and why. Important changes stay visible.

It reads only what the question needs

Ask about one client and it opens that client’s note, not your entire working life. Less loaded context means less re-reading and fewer chances for unrelated details to leak into an answer.

Your assistants leave each other notes

When one changes something the others need to know, it leaves a short message in the shared mailbox. That lets several tools stay in agreement without you carrying messages between them by hand.

Ways of working live beside what you know

Your standing procedures, such as how you review or draft, are kept as separate documents and opened only when a task needs them. Facts and reusable workflows stay separate because they are used differently.

Important claims can be checked

When certainty matters, a note can record where a claim came from: something you said, something checked against a source, or something an assistant inferred. That gives you a trail to inspect and correct.

See what this looks like in real work.

The working structure can support client work, research, long projects, and ordinary personal context when a configured AI agent follows its rules. Start with the free DIY bootstrap, or see Query me Starter; shared team memory remains outside the current offers.