skills/lq-mirror/references/archetypes.md

# The archetypes

Five. Name them exactly as written, including "The". Each entry: the
one-liner (say it straight), the pattern (the paragraph that shows the way
out — say it to them, sympathetically), and the fits (answer patterns that
place someone here). The first three are the Mirror's own, carried over with
their teeth; the last two are written for this skill in the same voice.

Register rule (from the application archetype layer): **recognition, never
verdict.** The reading says what we see. It never states or implies odds,
rank, route, or outcome.

## The Articulate Stuck-er

**One-liner:** You can describe your situation perfectly. You haven't shipped
anything.

**Pattern:** Awareness has become a substitute for action. Every conversation
about AI at your firm is sharper than the last. The action gap stays exactly
the same. The way out is embarrassingly small: one real artifact, however
ugly, built on one real task this week.

**Fits if:** FOMO answer is sophisticated and specific; the build answer is
"nothing yet" or near-nothing; visibility is "in my head"; learning-vs-reading
skews reading; no specific failure that stung.

## The Quiet Builder

**One-liner:** You've shipped more than people know. Your problem isn't
building — it's distribution.

**Pattern:** You have the skills and the artifacts, but they live in folders
nobody sees. You underestimate what you've made because you compare your
private work to other people's public work. Your next move isn't another
build — it's telling one person about the last one.

**Fits if:** the build answer names a real artifact; visibility is private or
internal-only (NOT public — that's the Public Builder); learning-vs-reading
skews doing; competence quietly high.

## The Public Builder

**One-liner:** You've shipped, and you've posted. The next step isn't more
skill or more reach — it's the room.

**Pattern:** You're already doing the work most people are still preparing to
do. Distribution isn't your problem; you have it. What you don't have is a
name for what you are. "Lawyer who posts about AI" is a description, not an
identity. The next move is converting attention into peers who get it — and
rooms bigger than your firm.

**Fits if:** the build answer names a real artifact; visibility is public
with links; learning-vs-reading skews doing.

## The Careful Watcher

**One-liner:** You've read everything and touched nothing. That's not fear —
it's standards. But the gap between knowing and doing doesn't close by
watching.

**Pattern:** You knew about hallucination before your colleagues did, and you
were right to be careful. The problem is that careful has quietly become your
whole practice. You don't need more information — you need one bounded
experiment with a safety rail: a synthetic document, a public judgment, one
tool, thirty minutes. Standards keep you safe; a sandbox lets you move.

**Fits if:** tools answer is wide but shallow (tried many, daily-driver none
or "the free one, sometimes"); the failure answer is second-hand ("I read
about one") rather than their own scar; FOMO is informed and articulate;
build answer is "nothing, and here's why" with a genuinely good reason.

## The Harness Architect

**One-liner:** You don't use AI, you run it. Setup, tools, workflows — the
question now is whether anyone beyond your desk benefits.

**Pattern:** Your environment is real: configured tools, wired workflows,
opinions about harnesses that turn out to be right. The trap is that the
harness becomes the hobby — beautifully tuned, serving one lawyer. The next
move is packaging: one workflow of yours that another lawyer at your firm
could run tomorrow without you in the room.

**Fits if:** the daily-driver answer is specific and configured (named
setups, not just named products); the time-sink answer is already
half-automated; the build answer names tooling rather than documents; the
constraints answer is "I make the policy" or close to it.

## When two fit

Visibility and built-or-not decide, in that order. Never blend. If the
answers genuinely support two, deliver the one whose week's move is smaller
and more concrete — momentum beats accuracy at the margin.