01
Repository awareness stays visible
The file tree is not hidden behind secondary screens. The product centers repo structure from the first meaningful workspace view.
Project-centered AI workspace
AI Tool Gateway is built around a project workspace instead of a scattered set of tool pages. Repository structure, file reading, model thoughts, and chat controls stay visually connected.
Workspace preview
The preview is intentionally separated from the hero so it can read like an actual interface instead of a compressed illustration.
Central presenter
def create_app():
app = Flask(...)
register_blueprints(app)
return app
Why this structure
The landing page should explain the product clearly, but it should also visually prepare the user for the kind of workspace they are entering.
01
The file tree is not hidden behind secondary screens. The product centers repo structure from the first meaningful workspace view.
02
Files and AI history share the same dominant reading surface, so the interface stays compact without feeling cramped.
03
Model thoughts, chat history, controls, and repo reading stay visually tied together rather than scattering across disconnected pages.
Product journey
The public surface stays minimal, while the signed-in flow grows into a denser, more capable project environment.
Landing, login, and account entry should be calm and obvious, with little visual noise.
Projects become the user-bound collection surface that leads into actual work instead of acting as an endpoint by themselves.
The project page becomes the real workspace: repo tree, central presenter, AI/chat history, and composer controls.
Design principles
The theme should feel deliberate and durable, with enough density for real work but enough spacing to remain clear.
Headers, panels, cards, toggles, and form controls share one theme language so page detail does not drift from the product shell.
The landing page should already feel compatible with account pages, project collections, and the richer workspace that follows.
Get started