#architecture

CMS Chronicle #12: The AI-Native Editor

748 commits in 11 days. We shipped a conversational CMS interface, upgraded the richtext editor to TipTap v3, built a media processing pipeline with Sharp and EXIF, added AI image analysis, and discovered a save bug that had been hiding since day one. [ICD test]

CMS Chronicle #11: Interactives, Media Adapters, and the Markdown Question

We built an entire Interactives engine, a pluggable Media Adapter system, GitHub-native media serving, and confronted the hardest architectural question yet: how do you embed rich interactive content inside markdown without breaking the separation of content and presentation?

CMS Chronicle #08: Multi-Site Admin One admin, many sites

A single cms-admin instance now manages multiple sites across multiple organizations. Supabase-inspired Sites Dashboard, org switcher, site-scoped everything — media, AI config, agents, settings. Plus the settings restructuring that made it possible.

The Age of the Agentic CMS Is Here. We're Building For It.

The CMS landscape is undergoing its biggest shift since WordPress. Here's what's happening — and why we built @webhouse/cms from scratch to meet this moment.

CMS Chronicle #06: Block Editor and Docker Block Editor & Docker

One image alignment bug that became a full block editor system, a broken upload architecture fixed three times, and two working deployments — local Docker and Fly.io.

CMS Chronicle #00: Why We're Building This Building This

A CMS is not a product decision. It's a philosophical one. We've shipped over 1000 websites. Here's why we're building a new opinionated CMS anyway — and what 30 years of lessons have taught us about what one should actually do.

CMS Chronicle #01: Building the Blog the Right Way

We built the Articles section of webhouse.dk this week. Twice. The first version hardcoded everything in TSX. The second version does it properly — content in CMS, rendering in code.

CMS Chronicle #02: Tags Cross-Collection

A CMS without tags is a filing cabinet without labels. We added tags to @webhouse/cms this week — as a first-class field type with cross-collection querying built in.

CMS Chronicle #04: SQLite Adapter The SQLite Adapter

Before Supabase, before PostgreSQL, we need an embedded database. SQLite is file-based, zero-server, and already sitting in package.json. Phase 3 started with archaeology — and three bugs that needed fixing.

CMS Chronicle #05: GitHub Adapter GitHub as a Database

GitHub is not just a git host. Every file in a repo is addressable, readable, and writable via REST API — with full history, branching, and access control built in for free. This is the GitOps CMS model: content as code, zero infrastructure.

@webhouse/cms