Skip to content
View ChestnutForest's full-sized avatar

Block or report ChestnutForest

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
ChestnutForest/README.md

Hi, I'm ChestnutForest 👋

Building tools that help people read, reproduce, and learn from AI research — mostly with Claude, FastAPI, and Next.js.

AI論文を読み解き、再現し、そこから学ぶための道具を作っています。 主に Claude・FastAPI・Next.js を使っています。


🔭 What I'm working on / いま取り組んでいること

Applying certification-level architecture patterns to a real product. I'm studying the Claude Certified Architect – Foundations (CCAR-F) curriculum, and applying what I learn directly to an application I'm building — then measuring how much of it I've actually applied.

認定試験レベルの設計の型を、実際のプロダクトに適用することに取り組んでいます。 Claude Certified Architect – Foundations(CCAR-F)で学んだ内容を、 開発中のアプリにそのまま適用し、どこまで適用できたかを計測して記録しています。

graph LR
    A["<b>Learn / 学ぶ</b><br/>CCAR-F skills<br/><br/><i>ccar-f-study-skills</i>"]
    B["<b>Apply / 実践する</b><br/>Real application<br/><br/><i>paper-repro-mvp</i>"]
    C["<b>docs/arch-guide/</b><br/>Coverage indicator<br/>適用率インジケーター"]

    A --> B
    B --> C
Loading

📦 Projects / プロジェクト

Repository What it is 概要
paper-repro-mvp A human-in-the-loop tool for reading and reproducing arXiv papers. FastAPI + Next.js. arXiv論文の読解〜再現実装を支援するツール(人間の承認を挟む設計)
ccar-f-study-skills Custom Claude Skills for CCAR-F exam preparation. CCAR-F試験対策のCustom Claude Skills集
processloop Process Dashboard (GPLv3) fork — Next.js port with en/ja i18n. Process Dashboard のフォーク。Next.js移植+日英対応
Deepware
antigravity-sandbox

🎯 Interests / 関心領域

  • Paper reproduction workflows — turning a paper into runnable, verified code (論文を、動いて検証できるコードに落とすまでの流れ)
  • Human-in-the-loop agent design — approval gates over full automation (全自動より、承認ゲートを挟む設計)
  • Structured output & reliability — JSON schemas, validation loops, provenance (構造化出力と信頼性 — スキーマ、検証ループ、出典の保持)
  • Knowledge capture — turning daily development into searchable, reusable logs (日々の開発を、検索・再利用できる記録に変えること)
  • Software process — PSP/TSP, measurement, and continuous improvement (ソフトウェアプロセス — 計測と継続的改善)

🛠️ Tech I use / 使っている技術

Backend Python · FastAPI · SQLAlchemy · PostgreSQL · Celery / Redis Frontend TypeScript · Next.js · React Tooling Claude Code · Docker · Git · pytest · Ruff


📊 A note on how I work / 進め方について

I keep a daily development log for each project, and I measure my own progress — including how much of what I've studied I've actually put into practice. It keeps learning and building connected instead of separate.

各プロジェクトで日次の開発ログを残し、進捗を自分で計測しています。 「学んだこと」と「実際に適用したこと」を切り離さないための工夫です。

Popular repositories Loading

  1. Deepware Deepware Public

  2. paper-repro-mvp paper-repro-mvp Public

    A human-in-the-loop tool for reading and reproducing arXiv papers (FastAPI + Next.js), built by applying CCAR-F architecture patterns

    Python

  3. processloop processloop Public

    Process Dashboard (GPLv3) fork — Next.js port with en/ja i18n

    TypeScript

  4. ccar-f-study-skills ccar-f-study-skills Public

    Custom Claude Skills for CCAR-F exam preparation

  5. antigravity-sandbox antigravity-sandbox Public

  6. ChestnutForest ChestnutForest Public