NichfuliOS APP BRAND Explore apps

Nichful

Main

HomeBrowse by purposeAbout NichfulSPEC driven quality designSupportPrivacy Policy

Apps

Maker's NotePMP Exam PrepAnthropic CertificationAWS CertificationMicrosoft CertificationMyFanClub

SPEC-DRIVEN DEVELOPMENT

Speed begins
with a clear spec.

Humans set the direction. Specialized agents build from the spec. Quality gates decide what is ready to ship.

See the spec engine ↘
INTENTI need exam prep
built for the new exam.
SPEC-2026.md

REQ-01 1,440 questions

REQ-02 5 languages

AC-03 Review before ship

01DESIGN
02BUILD
03VERIFY
Maker's Note portfolio screen in English built from the approved spec
RELEASE READY
2WEEKS

PMP ECO2026 version

1,440QUESTIONS

Structured in a focused build cycle

5LANGUAGES

Bringing learning to the world

2–3DAYS

From voice to improvement

01 / THE SPEC ENGINE

Turn one line of intent
into a spec no one has to guess.

Using PMP and CCA-F exam prep as examples, see how a learner's problem becomes a specification, verified content, and a working app.
LIVE SPECINTENT CAPTURED

START WITH WHY

Turn an English-only barrier
into understanding in your own language.

USER
I feel burdened by the English-based official materials even before I begin studying.
PROBLEM
Even if I only look at the correct answer, I can't understand why I got it wrong.
GOAL
Learn to understand in your native language and acquire the judgment skills necessary to pass the exam.

STRUCTURED REQUIREMENTS

Turn a user’s problem
into features and quality criteria.

  • LANGUAGEYou can learn by switching between your native language, English, and written language.
  • EXPLANATIONReasons for correct/incorrect answers for all questions/choices
  • EXAM FITCompliant with PMP ECO/CCA-F official domain ratio
  • LEARNINGReview of incorrect answers/weakness analysis/real-style mock exam

EXECUTABLE PLAN

Break a complete learning experience
into work that can be built and verified.

OFFICIAL SOURCEOrganize official materials CONTENT MAPAllocation to ECO/domain QUESTIONCreate questions and incorrect answers EXPLANATIONRationale for all options LOCALIZETranslated as a native language APP EXPERIENCEStudy/review/mock exam

PARALLEL AGENTS

Specialized agents
work in parallel, each going deep.

CONTENT AGENTOutline of the problem from official documents EXPLANATION AGENTThe basis for everything that is right and wrong LOCALIZE AGENTNative language and English written together APP AGENTStudy/review/mock exam

TRACE & TEST

Not just more questions—
quality that carries learning through.

SOURCE TRACECorrespondence with official guides, ECO, and AcademyPASS EXPLANATIONExplains not only the correct answer but also the reason for the incorrect answerPASS EXAM BALANCEReproduce the official domain and question ratioPASS LANGUAGE REVIEWTechnical terminology and naturalness as a native languagePASS HUMAN REVIEWHumans make the final decision on the content and whether or not it can be published.REVIEW

HUMAN DECISION

Before anything ships,
a human makes the final call.

BUILD 2026.07READY TO SHIP

02 / HUMAN × AGENT TEAM

Human direction.
An agent team
moves the work deeper.

HUMAN DIRECTIONPurpose/Priority/Final Judgment
MAIN ORCHESTRATORRead the SPEC and organize your work
PLANNINGRequirements arrangement
UIscreen design
BUILDimplementation
QAtest
LEGALRights/expression
LOCALIZEmultilingual

Go back and forth with suggestions, implementations, reviews, and modifications. We are a small development team that shares SPEC, rather than focusing entirely on AI.

03 / FOUR QUALITY GATES

Before release,
we stop at four quality gates.

Don't publish what you create quickly. Only when all four perspectives are aligned can we proceed to the decision to deliver.
GATE 01

SPEC
TRACEABILITY

Are the requirement ID, screen, data, and test perspective connected?

TRACE
GATE 02

IMPLEMENTATION
QUALITY

Are there any problems with state management, errors, billing, and impact on existing functions?

BUILD
GATE 03

CONTENT
& LEGAL

Were you able to confirm the basis, quality of incorrect answers, copyright, terminology, and translation?

CONTENT
GATE 04

RELEASE
READINESS

Have you prepared the actual equipment, examination, privacy, support, and improvement guideline?

SHIP
HUMAN DECISIONREADY TO SHIP

Humans decide whether to publish, revise, or suspend the work based on the AI's confirmation results.

04 / YOUR VOICE

Tell us
what should be better next.

The moment you have a problem, the function you wanted, the part that was difficult to understand. Please send us whatever you find out. After reading everything, think about the next improvement.

report the problemConsult how to useSend the features you want
Go to support/opinion box ↗

Whether you have a question that requires a reply or something you would like to share anonymously.

SPEC. BUILD. VERIFY. RELEASE.

Built from intent,
moved by the spec.

Explore apps ↗