mcpbeat

Fetch And Send Data

dotnet/fetch-and-send-data

Call APIs, load data into components, and handle the async lifecycle in Blazor. USE FOR fetching data from a backend, submitting data to an API, displaying loading/error states, registering HttpClient, building service abstractions for Auto/WebAssembly render modes. DO NOT USE for form validation (see collect-user-input), prerendering persistence (see support-prerendering), or project scaffolding (see create-blazor-project).

3k tokens
context cost
the whole folder, loaded on every use
1
files
instructions only
0
copies elsewhere
how many repositories repackaged it
4898
stars on the repo
on the repository, not the skill itself

Install

one command, takes just this skill from the repository
npx skills add https://github.com/dotnet/skills --skill fetch-and-send-data

The instruction itself

16 sections, as written by the author

Fetch and Send Data

Step 1 — Read AGENTS.md

Check Interactivity Mode and Scope:

| Mode | Data access |

|------|-------------|

| None (Static SSR) | Server-side: inject services/DbContext. Use [StreamRendering] for loading UX. |

| Server | Server-side: inject services/DbContext. Guard prerender with ??= + [PersistentState]. |

| WebAssembly | Browser-side: HttpClient only. No direct server access. |

| Auto | Both server and browser. Always go through an API. |

Step 2 — Register HttpClient

Only needed when calling external APIs from Server, or always for WebAssembly/Auto. Server components accessing their own database should inject DbContext or a service directly.

// Named client — requires Microsoft.Extensions.Http NuGet
builder.Services.AddHttpClient("CatalogAPI", client =>
{
    client.BaseAddress = new Uri("https://api.example.com/");
});

// Typed client
builder.Services.AddHttpClient<CatalogClient>(client =>
    client.BaseAddress = new Uri("https://api.example.com/"));

For WebAssembly/Auto with prerendering, register in both server and .Client Program.cs.

Step 3 — Fetch Data

Simple load

@page "/products"
@inject CatalogClient Catalog

@if (products is null)
{
    <p>Loading…</p>
}
else
{
    @foreach (var p in products)
    {
        <p>@p.Name — @p.Price.ToString("C")</p>
    }
}

@code {
    private Product[]? products;

    protected override async Task OnInitializedAsync()
    {
        products = await Catalog.GetProductsAsync();
    }
}

No error handling needed in the simplest case — wrap the component usage in <ErrorBoundary> at the parent/layout level to catch unhandled exceptions.

Static SSR — StreamRendering

Without [StreamRendering], the user sees nothing until OnInitializedAsync completes:

@attribute [StreamRendering]

Only affects Static SSR. No effect on interactive components.

Prerendering guard

Prerendering calls OnInitializedAsync twice. Skip the duplicate:

[PersistentState] private Product[]? products;

protected override async Task OnInitializedAsync()
{
    products ??= await Catalog.GetProductsAsync();
}

See the support-prerendering skill for details.

Step 4 — Handle Errors

Use <ErrorBoundary> as the default error strategy. It provides a consistent error experience across all components without any per-component catch logic. Wrap component usage at the layout or parent level:

<ErrorBoundary>
    <ChildContent>
        <ProductList />
    </ChildContent>
    <ErrorContent>
        <div class="alert alert-danger">Something went wrong. Please refresh.</div>
    </ErrorContent>
</ErrorBoundary>

Non-cancellation exceptions (HttpRequestException, etc.) propagate to ErrorBoundary automatically — no catch blocks needed in the component.

Cancellation is special

ComponentBase silently swallows all OperationCanceledException — both self-initiated (disposal, parameter change) and external (HttpClient timeout). ErrorBoundary never sees them. This means:

  • Self-cancellation → silently ignored. Correct behavior, no action needed.
  • External cancellation (timeout) → also silently swallowed. Component gets stuck in loading state. Usually acceptable — timeouts are rare.

When to add in-component error handling

Only add catch blocks when the component needs behavior ErrorBoundary can't provide — typically retries or timeout-specific messages. Even then, only catch what you need:

// Catch only external cancellation (timeouts) — everything else flows to ErrorBoundary
catch (OperationCanceledException ex) when (!cancellationToken.IsCancellationRequested)
{
    Logger.LogWarning(ex, "Request timed out for category {CategoryId}", CategoryId);
    error = "The request timed out. Please try again.";
}

If the component also needs to handle general errors with a retry button instead of letting ErrorBoundary take over:

catch (OperationCanceledException ex) when (!cancellationToken.IsCancellationRequested)
{
    Logger.LogWarning(ex, "Request timed out for category {CategoryId}", CategoryId);
    error = "The request timed out. Please try again.";
}
catch (Exception ex)
{
    Logger.LogError(ex, "Failed to load products for category {CategoryId}", CategoryId);
    error = "Unable to load products. Please try again.";
}

Rules

  • Never display exception.Message — it may contain PII, connection strings, or internal details. Use hardcoded user-friendly messages.
  • Always log through ILogger — the real exception goes to the logging pipeline.
  • Services must accept CancellationToken — pass it to every async call so work stops when the component cancels.

Step 5 — Parameter-Driven Reloading

When data depends on a route or query parameter that changes (e.g., navigating between /products/1 and /products/2), use OnParametersSetAsync with a guard to skip reloads for parameters that don't affect data.

Pattern: cancel-and-reload with stale data overlay

@page "/products/{CategoryId:int}"
@implements IAsyncDisposable
@inject ProductService ProductService
@inject ILogger<Products> Logger

@if (error is not null)
{
    <div class="alert alert-danger">
        <p>@error</p>
        <button @onclick="LoadAsync">Retry</button>
    </div>
}
else if (products is null)
{
    <p>Loading…</p>
}
else
{
    @if (isLoading)
    {
        <p><em>Refreshing…</em></p>
    }
    @foreach (var p in products)
    {
        <p>@p.Name — @p.Price.ToString("C")</p>
    }
}

@code {
    [Parameter] public int CategoryId { get; set; }
    [SupplyParameterFromQuery] public string? ViewMode { get; set; } // UI-only

    private CancellationTokenSource? cts;
    private int? loadedCategoryId;
    private List<Product>? products;
    private bool isLoading;
    private string? error;

    protected override async Task OnParametersSetAsync()
    {
        if (CategoryId == loadedCategoryId)
        {
            return; // Only ViewMode changed — no reload
        }

        loadedCategoryId = CategoryId;
        await LoadAsync();
    }

    private async Task LoadAsync()
    {
        if (cts is not null)
        {
            await cts.CancelAsync();
            cts.Dispose();
        }

        cts = new CancellationTokenSource();
        var cancellationToken = cts.Token; // Capture locally before await

        error = null;
        isLoading = true;

        try
        {
            var result = await ProductService.GetByCategoryAsync(CategoryId, cancellationToken);
            products = result;
        }
        catch (OperationCanceledException ex) when (!cancellationToken.IsCancellationRequested)
        {
            Logger.LogWarning(ex, "Timed out loading category {CategoryId}", CategoryId);
            error = "The request timed out. Please try again.";
        }
        finally
        {
            isLoading = false;
        }
    }

    public async ValueTask DisposeAsync()
    {
        if (cts is not null)
        {
            await cts.CancelAsync();
            cts.Dispose();
        }
    }
}

Key details:

  • Guard with tracked value: loadedCategoryId skips reloads when only UI parameters change.
  • Capture the token locally before the await — the CTS field may be replaced by a concurrent parameter change.
  • Don't null out products on subsequent loads — keep existing data visible with an isLoading overlay.
  • IAsyncDisposable cancels pending work when the user navigates away.

Step 6 — Send Data

var response = await http.PostAsJsonAsync("products", newProduct);
response.EnsureSuccessStatusCode();

var response = await http.PutAsJsonAsync($"products/{id}", updated);
response.EnsureSuccessStatusCode();

var response = await http.DeleteAsync($"products/{id}");
response.EnsureSuccessStatusCode();

Disable the submit button while saving to prevent duplicate requests. Show a saving indicator.

Step 7 — Service Abstraction for Auto or WebAssembly with Prerendering

When components run in both server and browser (Auto mode, or WebAssembly with prerendering), abstract data access behind an abstract base class:

public abstract class ProductServiceBase
{
    public abstract Task<Product[]> GetAllAsync(CancellationToken ct = default);
}

// Server — direct database access
public class ServerProductService(AppDbContext db) : ProductServiceBase
{
    public override async Task<Product[]> GetAllAsync(CancellationToken ct = default) =>
        await db.Products.ToArrayAsync(ct);
}

// Client — calls API
public class ClientProductService(HttpClient http) : ProductServiceBase
{
    public override async Task<Product[]> GetAllAsync(CancellationToken ct = default) =>
        await http.GetFromJsonAsync<Product[]>("api/products", ct) ?? [];
}

Register the appropriate implementation in each project's Program.cs. Components inject the abstract base class.

Don'ts

  • Don't call APIs in constructors — use OnInitializedAsync.
  • Don't use OnParametersSetAsync unless data depends on a changing parameter. Use OnInitializedAsync for initial loads.
  • Don't inject DbContext in WebAssembly/Auto components — no database in the browser.
  • Don't call your own server via HttpClient — inject the service directly.
  • Don't display exception.Message to users — PII risk. Log it, show a generic message.
  • Don't catch OperationCanceledException for self-cancellationComponentBase handles it.

How to use it

Copy the folder

Take dotnet/fetch-and-send-data from the repository into ~/.claude/skills for personal use, or into .claude/skills inside a project.

Check the name does not clash

The agent identifies a skill by the name field in its header. Two skills with the same name cannot sit side by side — one of them will be ignored.