Skip to content
Back to blog
Written by
Zinian
Published on
May 1, 2026
Read time
7 min read

Product Building

Lessons from Building KizunaIndex as a Public Index

A public index gets more useful when the content model is small, explicit, and easy to revise.

KizunaIndex started as a simple public surface for references, but the interesting part was never the index itself. It was the structure behind it.

The lesson

The small schema mattered more than the visual layer. Once the content primitives were clear, it became much easier to reason about:

  • What counts as a source
  • What counts as an entry
  • What needs review
  • What can be generated later

The shape that held up

Code
type Entry = { title: string; slug: string; summary: string; sourceLinks: string[]; status: "draft" | "ready" | "published"; };

Why public helps

Public systems force better boundaries. If a structure feels confusing when someone else sees it, it is usually confusing for the builder too. That made KizunaIndex useful as a design exercise, not just a publishing surface.

What I would keep

  1. Keep the model narrow.
  2. Keep the review step visible.
  3. Keep local editing first.
  4. Keep room for MDX when the content stabilizes.

The site works because the model stays understandable. That is not glamorous, but it scales better than cleverness.

Related work

Projects and case studies connected to this note.

Projects

Kizuna Index

A living index that turns scattered references into a maintainable publishing surface.

Product
AI Local Blog Editor

A local editing workflow for drafting, revising, and publishing technical posts with AI assistance.

AI Systems

Case studies

Turning a Static Index into a Maintainable AI-Assisted Publishing Workflow

A publishing system that keeps notes, sources, and draft output aligned without forcing a CMS.

Independent product build

Related posts

More notes close to this topic.

Designing Case Studies for Technical Portfolio Sites

Featured
Product Building
May 5, 20266 min readUpdated May 7, 2026

How to structure a case study so it proves judgment, not just output.

portfolio
case-study
writing
Read article

Building a Personal AI Workbench with Local Tools and Cloud Models

Featured
AI Engineering
May 8, 20268 min readUpdated May 10, 2026

A practical setup for drafting, testing, and shipping AI work without letting the workflow get tangled.

ai-agent
automation
codex
Read article

Shipping Small Web Products with Next.js, Vercel, and Supabase

Deployment
April 18, 20265 min readUpdated April 22, 2026

A small product stack works best when deployment, data, and UI stay easy to reason about.

nextjs
vercel
supabase
Read article

Next

Keep browsing the archive, or turn the questions in this essay into a concrete conversation.