From Resource to Web Dashboard - Automatically

Streamlit turns Python scripts into web apps (docs). Firestone leverages this to generate interactive dashboards automatically from your resource definitions.

You don't write any Streamlit code. Define your resource, and firestone creates a functional web UI with tables, forms, filters, and CRUD operations.

What Gets Generated

From this resource:

kind: book
schema:
  type: array
  key:
    name: book_id
  items:
    type: object
    properties:
      title:
        type: string
      author:
        type: string
      year:
        type: integer
        minimum: 1000
        maximum: 2100
      genre:
        type: string
        enum: [fiction, non-fiction, science, history]

Firestone generates a Streamlit app with:

List View

  • Table showing all books
  • Sortable columns
  • Search/filter capability
  • Pagination controls

Create Form

  • Input field for title (text input)
  • Input field for author (text input)
  • Number input for year (with min/max validation from schema)
  • Dropdown for genre (populated from enum)
  • Submit button that validates and creates

Detail View

  • Display single book information
  • Edit button to modify
  • Delete button to remove

Update Form

  • Pre-populated fields with current values
  • Same validation as create form
  • Save changes button

How It Works

Firestone analyzes your JSON Schema and generates appropriate Streamlit widgets:

Schema to Widget Mapping

  • type: string → st.text_input()
  • type: integer → st.number_input()
  • type: boolean → st.checkbox()
  • enum: [...] → st.selectbox()
  • type: string, format: date → st.date_input()

Validation Enforcement

  • minimum/maximum → number input bounds
  • minLength/maxLength → text input validation
  • pattern → regex validation
  • required → marked as required in UI

💡 One Definition, Four Outputs This same resource definition also generates OpenAPI (REST specs), AsyncAPI (WebSocket specs), and CLI (command-line tools).

Running the Generated Dashboard

# Generate the Streamlit app
firestone generate --resources book.yaml streamlit > app.py

# Run it
streamlit run app.py

Opens in browser at http://localhost:8501 with a fully functional dashboard.

Use Cases

Internal Tools Perfect for admin panels, data management tools, and internal dashboards without hiring frontend developers.

Prototyping Rapidly test your API design with a visual interface before building the production UI.

Data Exploration Give non-technical team members a way to view and manage resources without learning APIs or CLIs.

Demo & Testing Show stakeholders working functionality immediately after defining your resource.

Multi-Resource Dashboards

Generate one app with multiple resources:

firestone generate \
    --resources book.yaml,author.yaml,publisher.yaml \
    streamlit > app.py

Creates a multi-page Streamlit app:

  • Sidebar navigation between resources
  • Each resource gets its own page
  • Consistent UI across all resources

The Power of Consistency

Your Streamlit dashboard validates data the same way your OpenAPI spec does - they share the same JSON Schema. Add a validation rule to your resource:

properties:
  email:
    type: string
    pattern: "^[a-zA-Z0-9._%+-]+@[a-zA-Z0-9.-]+\\.[a-zA-Z]{2,}$"

Now validation works in:

  • OpenAPI spec (API requests rejected if invalid)
  • CLI (command fails if invalid)
  • Streamlit (form shows error if invalid)

One rule, three places it's enforced. Change it once, regenerate, it updates everywhere.

Next Steps

Ready to generate your first Streamlit dashboard?