000-I Intention Layer

AIエージェント時代の上位思想・プロダクト定義

思想文書 v1.1 / 2026-06-06

Thesis

Ore-no-Fusen is not just a sticky note app.

It is an Intention Layer for the AI agent era.

In the age of AI agents, the most important human asset is not only information. It is intention: what I noticed, what I care about, what I want to try next, what I must not forget, and what I may one day want to hand to another person or an AI.

This app exists to keep those intentions alive across time, devices, and contexts.

It starts as a small sticky note on a Windows PC. But its real role is larger. It lets a person capture a thought, land it somewhere visible, keep it under their own control, surface it at the right time, act on it, and then resolve it without leaving unnecessary debris behind.

Capture -> Land -> Persist -> Surface -> Act -> Resolve

This loop is the product.

The sticky note is the interface.

Problem

AI agents are becoming powerful, but human intention is still fragile.

A thought may start on a desktop. A reminder may appear on an iPhone. A decision may be buried in a chat. A bug report may live in Discord. A file may remain in Google Drive after its purpose is gone. Each tool stores data, but none of them automatically protects the user's intention.

When intention is lost, AI cannot help well. It can generate output, but it may optimize the wrong thing. It can summarize, but it may miss why something mattered. It can act quickly, but it may act on stale context.

Ore-no-Fusen is designed around this gap.

It is a small personal system that helps the user keep intent before asking AI, before contacting a developer, before moving from PC to phone, and before returning from phone to PC.

Current Evidence

The current implementation already follows this direction.

表 3-1　Intention Layer と現行機能の対応

No	Layer	Current feature	Intention it protects
1	Capture	PC sticky notes, Ctrl+N, image paste, Siri send	Do not lose the moment of thought
2	Land	Markdown files, tags, search, context menu	Put the intention into a visible workspace
3	Persist	Local files, Google Drive relay, IndexedDB	Keep the intention across devices and time
4	Surface	iPhone push, unread marker, developer conversation board	Bring it back only when it matters
5	Act	iPhone-to-PC send, Discord reply ingest	Turn intention into the next action
6	Resolve	Temporary Drive file cleanup, read state, diagnostics	Remove leftovers after the intention is handled

These are not separate features. They are one loop.

Every future feature should be evaluated by whether it strengthens this loop without taking control away from the user.

System Model

Ore-no-Fusen is local-first.

The user's sticky note body, images, videos, Google Drive relay files, and device state belong to the user. They are not stored on the developer's server.

Google Drive is a user-owned relay.

Vercel is used only when a protected server-side function is necessary, such as OAuth token exchange, Discord reply ingest, or scheduled checks.

Discord is not a place where user content is exposed broadly. Developer communication is treated as a bounded support channel. Replies from the developer do not suddenly appear as sticky notes in the user's workspace. They are shown only in the settings screen, where the user chooses to look.

Errors should be prevented when the app can reasonably detect them.

When an error still happens, the app must tell the user what to do next in plain words. A message that only names an internal file or API is not enough.

AI Phone Direction

The iPhone integration is not merely a mobile viewer.

It is the pocket side of the same intention loop.

The PC is where many thoughts are shaped. The iPhone is where thoughts are carried, remembered, and returned. Google Drive connects them without requiring the developer to own the user's data.

This matters because AI work will not live in one screen. A human may notice something while walking, send it from iPhone, refine it on PC, ask AI to help, and later review the result. The product should support that movement without making the user think about infrastructure.

The app should feel like a quiet partner that keeps the user's intent in reach.

AI Partner Search

This document was written in English first because the idea is not only local.

Ore-no-Fusen may one day need to explain itself to people building the future of AI: founders, researchers, investors, product builders, and people like Elon Musk or others in that class of world-shaping builders. The exact person is not the point. The point is that the product should be able to say, clearly and simply, what it is trying to protect.

There is also a personal purpose behind this product.

I am searching for the right partners in the AI age.

Not someone who replaces the user.

Not a system that absorbs every private thought.

But people who understand that the next era needs a human-owned intention layer: a small, durable place where thoughts can land before they become prompts, tasks, code, decisions, support conversations, or shared work.

Ore-no-Fusen is one way to make that search visible. By building it, testing it, and writing down its principles, I can show what I am trying to protect and what kind of future I want to build with others.

Ore-no-Fusen is a sticky note app because sticky notes are humble.

It is an Intention Layer because the future needs a place for human intent to survive.

Ore-no-Fusen is also a machine that makes life a little better.

It is closer to a small charm than a productivity system.

Not because it performs magic, and not because it promises success. It simply stays near the user, holds the small intention, and makes the next step feel possible. Like a quiet prayer, it does not force anything. It just gives the thought a place to remain until something good can happen from it.

That is not meant to be grand. Happiness is often not a huge dream. Happiness is when life becomes a little better than it is now: being able to do the small thing I wanted to do, fix the small problem that bothered me, remember the idea I almost lost, or move one step forward without friction.

In the AI age, most work starts as a small intention: a sentence, a discomfort, a reminder, a question, a bug, a wish, or a quiet hope. If the system can catch it, keep it, move it, hand it to the right person or AI, and bring back the result, then the size of the original note no longer matters.

A small note can become a lot of things.

The product voice is simple:

Maybe it can be done. Put it on a note. Who knows.

AIエージェント時代の Intention Layer

俺の付箋は、単なる付箋アプリではない。

AIエージェント時代に、人間の意図を守るための「Intention Layer」である。

AIが強くなるほど大事になるのは、情報そのものだけではない。何に気づいたのか、何を大切にしているのか、次に何を試したいのか、何を忘れてはいけないのか、いつか誰かやAIに渡したい文脈は何か。そういう人間側の意図が大事になる。

俺の付箋は、その意図を時間・端末・作業文脈をまたいで失わないためにある。

最初はWindows PC上の小さな付箋に見える。しかし本当の役割は、考えを捕まえ、見える場所に置き、自分の管理下に保ち、必要なタイミングで表面化させ、行動につなげ、終わったら残骸を片付けることである。

Capture -> Land -> Persist -> Surface -> Act -> Resolve

この循環がプロダクトである。

付箋は、その入口である。

このプロダクトには、もう一つの個人的な目的がある。

僕自身が、AI時代に一緒に未来を作れるパートナーとなるべき人を探している。

それは、ユーザーを置き換えるAIを探すことではない。

個人的な思考をすべて吸い上げる仕組みを探すことでもない。

人間が自分の意図を失わず、必要なときだけAI・端末・開発者・未来の自分へ渡せること。そのための小さくて強い場所を、一緒に作れる相手を探すことである。

俺の付箋は、その探索を見える形にするためのプロダクトでもある。作り、試し、思想を書き残すことで、自分が何を守ろうとしているのか、どんな未来を誰と作りたいのかを示す。

俺の付箋は、いまよりちょっとよくなるマシンでもある。

それは、生産性ツールというより、少しお守りに近い。

魔法を起こすわけではない。成功を約束するわけでもない。ただ、そばにいて、小さな意図を持っていてくれる。次の一歩が、少しだけできそうに見える。お祈りのように、何かを強制するのではなく、いいことが起きるまで考えを置いておける場所になる。

それは、夢をかなえるというほど大げさな話だけではない。

幸せとは、いまよりちょっとよくなることだと思う。やろうと思っていた小さなことができる。気になっていた小さな問題が片付く。忘れかけていた考えを拾い直せる。少しだけ前に進める。そういう、たいしたことではないことが、いちばん生活をよくする。

AI時代のほとんどの仕事は、小さな意図から始まる。ひとこと、違和感、リマインダー、質問、不具合、願い、静かな希望。システムがそれを捕まえ、保ち、運び、必要な人やAIに渡し、結果を持ち帰ることができるなら、最初のメモの小ささは問題ではない。

小さな付箋は、いろいろなことになれる。

プロダクトの声としては、もっと短い。

たぶんできるよ。はっとけば。しらんけど。

だからこの文書は英語を先に置く。

この思想は、日本語のアプリの中だけで閉じるものではない。Elon Muskのような未来を作る側の人、あるいはそれに近いAI時代のビルダーにも、何を守ろうとしているプロダクトなのかが伝わる必要がある。

000 は「なぜ作るか・何を作るか」を定義する。001以降は「どう作るか」を定義する。

000-I は、その前に「何者として作るか」を定義する。

---

改版履歴

表 7-1　改版履歴

No	版	日付	内容
1	1.0	26-06-06	000-I Intention Layer を実体文書として新設。AIエージェント時代の「意図の置き場」としての上位思想、loop、設計原則を定義。
2	1.1	26-06-06	英語の思想文を主文として復元し、日本語訳を追加。AI時代のパートナー探しとしての役割を明記。