From d7a8400392ffa1c92e6529b8a9536291e08e37d0 Mon Sep 17 00:00:00 2001 From: Tomasz Zajac Date: Fri, 25 Sep 2026 19:42:00 +0200 Subject: [PATCH] feat(website): trim landing page copy Cut visible text on the landing page from ~760 to ~310 words. Every tile and feature is now one short line; structure, icons and links are unchanged. Details stay in the manual, which the tiles link to. Co-Authored-By: Claude Opus 5.5 --- website/index.html | 36 +++++++++++++++++------------------- 1 file changed, 17 insertions(+), 19 deletions(-) diff --git a/website/index.html b/website/index.html index cb3a2bf7..b09efc59 100644 --- a/website/index.html +++ b/website/index.html @@ -40,9 +40,7 @@

Architecture that lives here,
and holds whatever builds from it.

- Diagrams go stale because the drawing is the document, - but the model doesn't stop at the diagram. Radical.Tools keeps one model, - modelled by you or by AI, and turns it into the harness whatever builds from it has to run inside. + One model, built by you or by AI, that becomes the harness for whatever builds from it.

Open Studio in the browser @@ -56,17 +54,17 @@

Architecture that lives here,
and holds whatever b
Modelled by either - Draw it yourself in Studio, chat it into being with the AI assistant, or let a coding agent edit the .radical file directly in your repo. The model doesn't care who's architecting today. + Draw it in Studio, chat it with AI, or let an agent edit the .radical file.
Read as context - Before it writes a line, a coding agent reads the model: the boundaries the metamodel allows, the decisions already recorded in ADRs, the requirements it has to satisfy. + Agents read boundaries, ADRs and requirements before writing code.
Checked against - Fitness functions, ADRs and EARS requirements aren't written after the fact. They're what catches the code, human or AI, when it drifts from what the model says. + Fitness functions and requirements catch code that drifts from the model.

@@ -76,22 +74,22 @@

Architecture that lives here,
and holds whatever b Metamodel01 - The rules of your architecture: element and relation types, what may contain what, which properties each type carries. Three presets ship built in — C4, C4 + DDD, C4 + DDD + Governance — and a visual editor lets you extend or replace them. + The rules: types, relations, containment. C4, DDD and Governance presets, fully editable. Model02 - One graph of elements and relations, validated against the metamodel as you edit — wrong parent, wrong relation or too many of a kind are refused, not silently drawn. Add an element once and it exists in every view. + One validated graph. Add an element once, see it in every view. Views03 - Six readings of the same model: Structure diagrams, numbered Flow sequences, a drillable Hierarchy treemap, an editable Table, a dependency Matrix and a Wiki of element pages. Each view keeps its own scope, layout and camera. + Structure, Flow, Hierarchy, Table, Matrix and Wiki — six readings of one model. Milestones04 - Freeze the model under a name at any point. Compare two milestones and see what was added, removed or changed as ghosts on the canvas; restore any of them. Decisions get a timeline. + Freeze, compare and restore named snapshots of the model. @@ -101,31 +99,31 @@

Also in the box

  • - Governance as first-class elements. ADRs, fitness functions and EARS requirements are model elements, linked to what they constrain, implement, satisfy or supersede. Type a requirement as a sentence — “When the user clicks save, the system shall persist the document” — and it is parsed into trigger, condition and action. + Governance built in. ADRs, fitness functions and EARS requirements as linked model elements.
  • - AI that edits the model, not the picture. Ask what depends on what, find gaps, or describe a change in plain language. The answer comes back as a patch that is validated against the metamodel, applied to the model and undoable. Claude, OpenAI, Gemini or local Ollama — your key, called directly, nothing proxied. + AI that edits the model. Validated, undoable patches. Claude, OpenAI, Gemini or Ollama — your key, nothing proxied.
  • - Concept Hub. 120+ curated requirements, ADRs, fitness functions, patterns and blueprints, rendered by the same engine as Studio. Fill in a template’s parameters, add it to your model, reconfigure later. + Concept Hub. 120+ ready-made requirements, ADRs, patterns and blueprints.
  • - Layout that earns its name. Smart Layout runs layered (ELK), force-directed and simulated-annealing candidates, scores them for crossings and overlap, and keeps the best. Live physics keeps things tidy while you drag; a tree mode handles deep containment. + Smart Layout. Tries several layout engines and keeps the cleanest result.
  • - Three perspectives, one model. Designer edits. Viewer explores read-only — drag, collapse, re-layout without touching the saved model. Presenter turns views into slides that capture scope, layout and camera; flow views replay step by step. + Design, view, present. Edit, explore read-only or turn views into slides.
  • - Plain files, no lock-in. Save as a single .radical JSON file, or as a folder of Markdown — one .md per element with YAML front-matter, hierarchy as directories, layout in a sidecar. Diff it in git, edit it in any editor, script it. Autosaved either way. + Plain files, no lock-in. One .radical JSON file or a folder of Markdown — git-friendly.
  • - Desktop, browser or VS Code — same code. The Electron app opens local files and folders and reloads a file you launched it with when it changes on disk. The web build needs no install and can still work on a local folder. The VS Code extension embeds the full editor for .radical files. + Desktop, browser or VS Code. Same editor everywhere.
@@ -134,7 +132,7 @@

Also in the box

Start from the Hub

-

Don’t model from a blank canvas. Open a concept as a diagram, wiki page or table — exactly as it will look in your model — collect the ones you want and add them to Studio in one click.

+

Skip the blank canvas — pick concepts and add them to Studio in one click.

Blueprints @@ -149,7 +147,7 @@

Start from the Hub

Run it locally

-

Node.js ≥ 18. The desktop app adds local folders, file watching and the Markdown-folder format with no browser permissions — or skip the install and use studio.radical.tools.

+

Node.js ≥ 18 — or skip the install and use studio.radical.tools.

GitHub → Releases →