Creator of the LLMO Framework. I design harnesses that automate dev, ops, and marketing with Claude Code — then publish what actually happened, with numbers.
400K+ article views (Qiita + Zenn) · 40+ books (Kindle + Zenn) · 4 research papers
Building AI-search-ready content? Start with the LLMO Framework.
If this work has saved you time, you can support its continued maintenance through GitHub Sponsors or Ko-fi.
| LLMO Framework | The open methodology for LLM Optimization (AEO + GEO) — getting your content found and cited by AI search |
| llmo-checker | Lighthouse-style CLI that scores how AI-retrievable a URL is: llms.txt, JSON-LD, robots policy, semantic structure → 0–100 LLMO Score |
A focused set of tools I use daily to keep Claude Code sessions productive across long work. Follow one, you'll want the others.
| compact-ops | Preserve session state across /compact — no lost context, no manual re-briefing |
| claude-shift | Switch between multiple Claude Code accounts safely; observe 5-hour / weekly usage across them |
| hook-chain-lens | Diagnose your Claude Code hooks: see the merged fire order across user / project / plugin scopes, catch dead matchers and missing script paths |
More: open-llmo (research initiative) · voice-clone (3-sec voice cloning on Qwen3-TTS) · speech-habit-lens (speech analyzer) · domain-pre-flight (domain checks) · persona-hub (persona SDK)
New products ship weekly: kenimoto.dev → Products
- When Free Executors Cost More: The Free-Executor Paradox in Iterative LLM Code-Repair Loops — DOI:10.5281/zenodo.20978074 · code
- Excess Vocabulary in Japanese AI-Generated Text — DOI:10.5281/zenodo.19233934 · code
- AI Text Slop: Stylistic Convergence Across Six LLMs — DOI:10.5281/zenodo.19173035 · code
- AI Blue: Color Recognition Bias in Vision-Language Models — DOI:10.5281/zenodo.19159702 · code
 Harness Engineering |
 Context Engineering |
 Claude Code |
 LLMO |
 Knowledge Graph |
 300ms Threshold |
Full catalog (EN / JA / PT / ES): kenimoto.dev → Publications
|
I stopped making Claude Code draw my diagrams and started drawing with it — 2026-08-04 Claude Code: The Operator's Field Guide — 2026-08-04 llms.txt vs Claude Skills manifest: I tested 5 AI engines to see which file they actually read — 2026-08-03 Chat Models Are Born in the Loss Mask — Reading nanochat's SFT — 2026-08-02 More on kenimoto.dev |
公開repoに実名が漏れる経路はdiffだけじゃなかった。gitフックで塞いだ7つの観点 — 2026-08-04 AIに引用される文章は明快さで決まる:5つの構造で設計する — 2026-08-03 Qwen 35Bの品質を7つの質問で採点したら、GPT-4に勝てるのは3領域だけだった — 2026-08-02 会議の合意を4パターンに分類したら、3つは合意ではなかった — 2026-08-01 感情をKGにしたら、AIの共感応答が7パターンに分岐した — 2026-07-31 More on zenn.dev/kenimo49 |
チャットモデルは「損失マスク」で生まれる ― nanochatの8,159行でSFTを読む — 2026-08-02 AI要約が「詐欺ではありません」と答える仕組み、踏み台はあなたのサイトの検索窓 — 2026-07-28 Claude Code Max $200/月×8ヶ月を私のタスク別に分解 — 4割はローカルLLMで代替できました — 2026-07-27 Claude Codeの/compactは、直前のtool結果を7割捨てます — 実測でわかった保持境界 — 2026-07-20 Qwen 35Bは平均42Wで動く — RTX 4070の消費電力プロファイルを推論中に4パターン実測 — 2026-07-20 More on qiita.com/kenimo49 |
If my tools or research helped you, you can sponsor my open-source work.
This README rebuilds itself daily via GitHub Actions, pulling from kenimoto.dev/ai/publications.md (the AI-readable feed), Zenn, and Qiita. Last refreshed: 2026-08-05 06:27 JST