Léandre Site Manifest

Why it exists

Before this site, the public surface was too thin for the actual range of the work.

It did not make it easy to centralize ongoing projects, connect them to notes, or share a more faithful picture of how things are built.

This project exists to fix that.

Not with a louder portfolio, a visual CV, or a content machine dressed up as identity.

Instead, with a public system that can hold projects, decisions, and technical choices in one place without flattening them.

It also works as a living prototype. Small enough to move quickly. Serious enough to prove a standard.

What the project proves

This site is a public proof of how I build: direction, editorial structure, and technical execution held together in one readable system.

The goal is not to show every capability separately.

It is to make the logic behind the work legible:

• a minimal public structure instead of more navigation
• projects and notes linked as one network
• technical choices that support clarity rather than performance theater
• a visual language that stays restrained, precise, and authored

The first move was documentation

The project did not start with components.

It started with a large documentation pass.

Product brief, editorial rules, design direction, content contracts, technical strategy, infrastructure constraints, risk notes, deployment doctrine: each layer was pushed until the concept became hard enough to build without drifting.

That mattered because the site was never meant to be a loose personal homepage.

It had to become a coherent system before becoming an interface.

What is already real

Under the current public surface, the build is already doing real work:

• a local MDX content model with strict metadata and explicit registries
• linked posts and project pages that stay statically knowable at build time
• a minimal public architecture centered on Home and Blog
• deterministic metadata outputs like sitemap.xml, feed.xml, and canonical metadata
• a static artifact designed to stay portable and easy to host
• a deployment posture shaped for Coolify and Cloudflare on a shared VM

There is still no CMS, and that is deliberate.

I already pilot the whole site through local MDX and Codex CLI: structure, copy, rewrites, metadata, and internal links all move faster there than they would in an admin layer. On a site with no products and almost no image workflow, a CMS would add more friction than value.

The site reads more like a compact publishing system than a dressed-up profile page because of that.

Codex is part of the build, not the direction

Codex is named here because it materially changed the speed and shape of the project.

The concept, the target stack, the constraints, and the editorial standard were mine from the start. Codex became the execution layer I could pressure-test against, challenge, and use to move faster across documentation, prototyping, implementation, and rewriting.

That changes a project like this.

The site can iterate at the speed of the idea instead of waiting for a longer handoff chain between strategy, writing, design, and code. For a public prototype with relatively low operational risk, that compression is especially useful.

Why Pretext stays targeted

Pretext is one of the modern pieces I specifically wanted to explore here.

Not as a theme. Not as a gimmick.

As a signature layer.

It belongs on the surfaces where better line control, measured width, and typographic rhythm actually improve perception. Everywhere else, plain CSS should keep winning.

That guardrail keeps the site exact instead of theatrical.

Current state

This project is live, useful, and still moving.

The foundation already exists. The editorial system exists. The stack is in place. The public shape is readable.

What remains in motion is refinement:

• stronger project pages as the corpus grows
• sharper notes connected to real work
• more precise signature surfaces where they earn their place
• repeated rewrites as new tools, better patterns, and better proof appear

That is part of the point.

This site was never meant to be finished once. It was meant to stay close to the work as the work changes.