Use the language, not a lecture.
Games begin with a language behavior: meaning, listening, retrieval, register, response, or production.
Language practice through games
Short, focused games that make language practice worth returning to. Korean comes first, with English guidance here and a separate Japanese-guided version.
Early product prototype • direct task feedback • no proficiency or learning-outcome claim
Games begin with a language behavior: meaning, listening, retrieval, register, response, or production.
Accuracy, latency, assistance, direction, and context stay attached to the task that produced them.
Foundation through Advanced changes the linguistic demand—not the maturity of the interface, situations, or player.
Playable language prototype
Signal Shift samples meaning, adult functional context, and social nuance. Multiple-choice recognition is diagnostic practice only: it is not productive mastery, scheduled-memory evidence, JLPT/CEFR scoring, or validated proficiency.
READY / KOREAN
Choose a lane, then begin.
Press start to begin.
Diagnostic-recognition demo only. First-response latency and assistance describe this task, not memory, fluency, or learning gains. If measurement is allowed, Nakimaru records a session summary but not individual answer choices.
The planned language system
Nakimaru is being designed around short daily language sessions, varied game mechanics, explicit sophistication bands, and progress views that say exactly what was measured.
A compact sequence can change mechanic while staying inside one language lane and one coherent practice history.
Planned history keeps direct measures separate so recognition cannot quietly masquerade as production.
Internal bands organize practice complexity without claiming JLPT, CEFR, school grade, fluency, or validated proficiency.
04 / WHERE THE GAMES CAN GO
Evidence discipline
Nakimaru can report what happened inside a language task: first-attempt accuracy, response latency, assistance, direction, context/register performance, and whether the task required recognition or production. Broader learning or proficiency claims require evidence the product does not yet possess.
01 First attempts remain distinct from assisted retries.
02 Recognition does not masquerade as productive mastery.
03 Internal bands do not masquerade as JLPT, CEFR, or fluency levels.
04 Learning-outcome claims require evidence before marketing.
Planned access model
A Nakimaru Labs membership is planned with Korean first, English and Japanese guidance, and more languages to follow.
NAKIMARU LABS
NAKIMARU / LABS