Quick start¶
This guide starts ShinyHub locally, validates an application, previews the deployment, and opens the routed result.
Start the server¶
Complete the first-run administrator setup and leave the server running. The
default address is http://localhost:8080.
Create a small Python Shiny app¶
Create app.py:
from shiny import App, ui
app_ui = ui.page_fluid(
ui.h2("Hello from ShinyHub"),
ui.p("A production-shaped deployment in a few commands."),
)
app = App(app_ui, server=None)
Create requirements.txt:
Validate locally¶
doctor reports all bundle and runtime problems in one pass. run --check
boots the app through the production-shaped /app/<slug>/ route and exits after
readiness succeeds.
Connect and deploy¶
shinyhub connect http://localhost:8080 --name local
shinyhub doctor .
shinyhub plan . --slug hello
shinyhub deploy . --slug hello --open
plan is read-only. It shows the exact archive, inferred launch command,
manifest changes, permissions, and start/stop effects before deploy changes
the server.
Continue¶
- Add names, limits, scaling, and lifecycle settings in the application manifest.
- Configure environment variables and encrypted secrets.
- Move persistent files into the application data directory.
- Use a fleet manifest to reconcile several applications.