I
Enter the library
A warm steampunk reading room. A brass star atlas on the central table. The learner is greeted, not graded.
A New Vision · Coming to Strata Mundo
The Star Atlas
Library
A thorough K-2 math mastery assessment, reimagined as an immersive steampunk library. Every standard is a star. Every probe reads the learner’s reasoning, not just their answer. Every result becomes a passport into the larger math universe.
78 standards · 31 constellations · K through grade 2 + the bridge to multiplication
The Promise
Tell a homeschool guide or microschool teacher exactly what their learner understands, what they don’t, and what to do about it— in fifteen minutes, without a test that feels like a test.
The Voyage
Five steps from arrival to passport
- II
Pick a probe card
The atlas highlights the next standard. The learner walks to the matching shelf and lifts a brass-trimmed card.
- III
Reason inside a probe
A 60-to-90-second instrumented mechanic. Telemetry watches every placement, every reset — the trajectory, not just the answer.
- IV
The atlas updates
The probed star turns gold. The voyage continues along the prerequisite chain. The session ends naturally when mastery thins out.
- V
Carry the passport
A mastery passport drops into the learner’s pocket. When they enter Math Games Builder next, the galaxy already knows them.
The Atlas
78 stars · 31 constellations
Every Common Core standard from kindergarten through second grade, plus the bridge into third-grade multiplication. Tap any star to see how we will measure that piece of mathematics — or try a working probe demo for the brightest star, K.CC.A.1.
KindergartenGrade 1Grade 2Bridge to grade 3
The Engine
Why this is hard to fool
Telemetry, not multiple choice
The trajectory tells the truth
Most diagnostics ask 'right or wrong.' We watch how the learner reasons. A child who tries one approach, resets, and tries another shows mastery. A child who clicks fast and wrong shows guessing. A child who clicks fast and right shows fluency — never penalized.
A four-stage AI ladder
Every mastery claim is checked
When the assessment finishes, four AI agents review the verdict in sequence: a Critic asks if the call is defensible from the data; a Shortcut Adversary tries to prove the learner faked it. Both run in fast and deep passes. If anyone disagrees, the verdict downgrades — never upgrades.
Built on research, not opinion
Every probe is a research artifact
Before any probe ships, three more agents review it: Mr. Chesure checks that it actually tests the standard it claims to. The Equity Reviewer checks for English-Learner accessibility and stereotype-threat traps. The External Reviewer hunts for what our other agents missed.
Carries forward, never gates
A passport, not a permission slip
The result is a JSON passport — a small living map of what the learner knows. It flows into Math Games Builder, lights up the moons they've mastered, opens what's next. It never locks anything. The guide always drives. The kid always plays.
After the Atlas
The fun door comes first
When a standard is flagged, the learner sees a game first — built by older learners in Math Games Builder. If they can win it, that’s real evidence of learning. If they can’t, the difficulty becomes the motivation to try the curated practice that the guide picks from a vetted library.
I
The fun door
A Math Games Builder game
A peer-built game targeting that standard. No gating — play it whenever. Winning is real evidence. Losing is real motivation.
II
The curated library
Khan, PhET, MLC, Montessori
A vetted menu of trusted resources tagged by misconception and modality. The guide picks. Concrete, then representational, then abstract.
III
Community contributions
Activities from other guides
Other guides submit what worked for them. AI vets, Equity Reviewer screens, a human approves. The library grows with the network.
See what we have today
The current Strata Mundo is a working diagnostic for grade 3-4 fractions — the proof of concept that the trajectory-as-truth idea works. The K-2 atlas above is what comes next.