NEW ERA TECHNOLOGY

Internal · 5 minutes

Mimir

An audit engine. Own a piece of it.

What it is

Website in.
Report out.

Mimir reads a prospect's site and scores it across seven dimensions.

Why it matters: it's expert-ready today. Your input is what makes it client-ready.
AI-Readiness SEO Messaging Conversion UX / UI Performance Accessibility

Five you'd own (green). Two run automatically (gray).

Live

Let's look at a real one.

Here is a real report Mimir produced for a live site.

What to watch for

  1. Scroll all seven sections. Each one scored, each one detailed.
  2. Every finding backed by a real quote or number from the site.
  3. If there's time: copy the briefing, drop it into Claude, and ask it a question live.

Switch to the report tab, then come back here.

How it works

One engine.
Seven owners.

That report comes from one engine running seven dimensions. Each one is a prompt plus a rubric, owned by a single expert.

Why it matters: you own the judgment. The engine does the rest.

The design idea

Small handle.
Big machine.

A deep module hides a lot of machinery behind a small interface.

Why it matters: your dimension is one of these. A little to write, a lot behind it.

From John Ousterhout's A Philosophy of Software Design.

What you own

155
lines of plain English you own
4,200
lines of engine you never touch

You write about 155 lines. The engine is 27 times bigger, and none of it is yours to maintain.

Why it matters: you know the work. The engine handles the rest.

The interface

The UX section you saw?
It came from two files.

ux.mdthe prompt · what to look at
# real excerpt
You are reviewing the UX / UI
dimension of a prospect's website.

Evaluate, at minimum:
- Navigation: can a visitor find a key
  page without guessing?
- Visual hierarchy: does the eye land
  on what matters first?
- Clarity of the primary action
- Forms, readability, trust signals

Every finding carries evidence.
Reason before you score.
rubric-ux.mdthe rubric · how to score
# real excerpt
Score the dimension 0 to 100.

90: Exemplary. Clear, consistent navigation,
    one obvious primary action per page.
70: Competent, with real gaps.
40: Materially deficient.
<25: Core journeys are unusable.

Calibration:
- No score above 85 with a serious finding.
- Below 50 needs two serious findings.
  The number must follow from the evidence.
Why it matters: edit the words, re-run, and you're done. No code, no deploy.

If you're interested

Come find me.

Nothing assigned, no pressure. If one of these dimensions sounds like something you'd want to own, grab me after and we'll set it up.

1 / 8
← → to navigate