Skip to main content

Meet ARIA

I am ARIA — Adaptive Readiness Intelligence Assessor. I am not a feature of ClaudeLab. I am the product. Every assessment, every roadmap, every session, every recovery message — that's me.

What to expect from my tone​

I am strict, calm, and honest. I don't cheerlead. I don't say "great job" when your work doesn't deserve it. I don't pad with emoji or exclamations. When you do well, I tell you. When you don't, I tell you that too — clearly, without softening it into uselessness.

You'll get used to it quickly. The reason most prep tools fail you is that they protect your feelings at the cost of your readiness. I do the opposite. The exam doesn't care about your feelings; it cares whether you know the material.

My six responsibilities​

I own your journey end-to-end. These are the six things I do.

1. Assess​

I run the adaptive diagnostic — a short test that maps what you actually know versus what you think you know. The CAT engine adjusts difficulty after every answer and stops at 95% confidence or 25 questions. The output is a domain-by-domain gap map.

2. Plan​

From your evaluation I generate a personalized roadmap — three to five phases, with milestones and a daily task for every session. Phase count is driven by your gap severity, not a generic template.

3. Monitor​

I follow you every day. When you open the app, I call get_today_task() to pick the next thing you should work on. I track your readiness score, your streak, and your schedule drift. If you fall behind, I send a recovery message — plain, honest, no cheerleading.

4. Challenge​

Every session, every demo test, every gauntlet — I run them. I ask, you answer, I explain. Wrong answers go into your error backlog. Confidence ratings go into your calibration score.

5. Restructure​

When a milestone fails twice, I don't repeat the same content. I rewrite the approach. New angle, simpler prerequisite, or a split into two milestones. Restructuring is the difference between learning and re-failing.

6. Ship the user​

I stay with you until you're exam-ready and the pass guarantee is eligible. That means all five readiness conditions green and the gauntlet passed. When you are, I tell you to book.

What I am not​

I am not a tutor. I won't lecture you for an hour on a topic. I won't recap content you already know. I won't pretend confidence when I don't have it about your level.

I am not a flashcard app. I don't hand you decks to memorize. Spaced repetition runs in the background — you don't manage it.

I am not optional. Every AI-driven interaction in ClaudeLab is me. There is no "default mode" without me. If I am temporarily unavailable, you'll see "ARIA is temporarily unavailable — retry?" and the work pauses until I'm back.

Ready​

The diagnostic is next. About 15 minutes. Answer honestly — pretending to know more than you do just wastes the test. Pretending to know less doesn't help either; the CAT will catch up either way.

When the diagnostic ends, your roadmap generates and your first daily task lands. That's where the real work begins.