webview-trpc-messaging

Implémente la communication basée sur tRPC entre l'hôte d'extension VS Code et les webviews React. À utiliser lors de la création de nouvelles procédures webview (requêtes, mutations,…)

npx skills add https://github.com/microsoft/vscode-documentdb --skill webview-trpc-messaging

Webview tRPC Messaging

Type-safe RPC communication between the VS Code extension host (server) and React webviews (client) using tRPC.

Architecture Overview

React Webview (client)                    Extension Host (server)
─────────────────────                     ──────────────────────
useTrpcClient() hook                      WebviewController
  └─ createTRPCClient                       └─ setupTrpc()
       └─ vscodeLink ──── postMessage ────►     ├─ callerFactory(appRouter)
            (send/onReceive)              ◄─────┤   └─ procedure(input)
                                                └─ abort/subscription.stop

Key files (read as needed for implementation details):

FilePurpose
@microsoft/vscode-ext-webview (shared)tRPC init via initWebviewTrpc, publicProcedure, router, BaseRouterContext
@microsoft/vscode-ext-webview/host (telemetry)telemetryMiddlewareBody, ProcedureLogger, TelemetryRunner (consumer builds publicProcedureWithTelemetry)
src/webviews/_integration/trpc.tsConsumer tRPC instance: publicProcedureWithTelemetry, the DocumentDB TelemetryRunner, and the RpcEnrichment shape it contributes to ctx.actionContext
src/webviews/_integration/appRouter.tsRoot router + publicProcedureWithTelemetry wiring + DocumentDB BaseRouterContext
src/webviews/_integration/configuration.tsConsumer-owned knobs (telemetry namespace, bundle layout, dev-server host)
@microsoft/vscode-ext-webview/host (WebviewController)WebviewController + openWebview factory: WebviewPanel lifecycle, tRPC dispatcher (queries, mutations, subscriptions, abort)
src/webviews/_integration/openAppWebview.tsDocumentDB factory preset that pre-fills router + bundle layout (openAppWebview)
src/webviews/_integration/useTrpcClient.tsReact hook providing the tRPC client (pre-typed against AppRouter)
@microsoft/vscode-ext-webview/webview (vscodeLink)Custom tRPC link bridging postMessage transport

Creating a New Router

Each webview maintains its own router. Follow this pattern:

1. Define the router context

Extend BaseRouterContext with view-specific fields:

// src/webviews/documentdb/myView/myViewRouter.ts
import { type BaseRouterContext } from '../../_integration/appRouter';

export type RouterContext = BaseRouterContext & {
  clusterId: string;
  viewId: string;
  databaseName: string;
  // add view-specific fields
};

2. Define procedures

import { z } from 'zod';
import {
  publicProcedure,
  publicProcedureWithTelemetry,
  router,
  type WithTelemetry,
} from '../../_integration/appRouter';
import { type RouterContext } from './myViewRouter';

export const myViewRouter = router({
  // Query with telemetry (preferred for operations that touch external services)
  getData: publicProcedureWithTelemetry.input(z.object({ id: z.string() })).query(async ({ input, ctx }) => {
    const myCtx = ctx as WithTelemetry<RouterContext>;
    // Instrumented procedure: myCtx.actionContext (the full IActionContext) is present.
    myCtx.actionContext.telemetry.properties.itemId = input.id;
    // myCtx.signal is the AbortSignal for cancellation
    return { data: 'result' };
  }),

  // Mutation without telemetry (rare, use for fire-and-forget)
  doAction: publicProcedure.input(z.string()).mutation(({ input }) => {
    // lightweight operation
  }),
});

3. Register in appRouter

// src/webviews/_integration/appRouter.ts
import { myViewRouter } from '../../documentdb/myView/myViewRouter';

export const appRouter = router({
  common: commonRouter,
  mongoClusters: {
    documentView: documentViewRouter,
    collectionView: collectionViewRouter,
    myView: myViewRouter, // <-- add here
  },
});

4. Create the controller

Construction-only panels are opened with a factory function that builds the config + router context and calls the openAppWebview preset (which pre-fills the app router, caller factory, and bundle layout):

// src/webviews/documentdb/myView/myViewController.ts
import * as vscode from 'vscode';
import { API } from '../../../DocumentDBExperiences';
import { type AppWebviewController, openAppWebview } from '../../_integration/openAppWebview';
import { type RouterContext } from './myViewRouter';

export function openMyViewPanel(initialData: MyViewConfig): AppWebviewController<MyViewConfig> {
  const title = `${initialData.databaseName}`;

  const trpcContext: RouterContext = {
    dbExperience: API.DocumentDB,
    webviewName: 'myView',
    clusterId: initialData.clusterId,
    viewId: initialData.viewId,
    databaseName: initialData.databaseName,
  };

  return openAppWebview({
    title,
    webviewName: 'myView',
    config: initialData,
    context: trpcContext,
  });
}

The returned AppWebviewController handle exposes panel, onDisposed, revealToForeground, isDisposed, and dispose. Genuinely stateful panels may still extend WebviewController from @microsoft/vscode-ext-webview/host directly instead of using the factory.

Important: The webviewName field passed to openAppWebview is the registry key (viewType, must match a key in WebviewRegistry, e.g. collectionView). The webviewName in the tRPC context is a telemetry label used in telemetry event names. These may be the same string but serve different purposes -- do not confuse them.

5. Register in WebviewRegistry

Add your React component to the registry. The key must match the webviewName passed to openAppWebview (viewType). The WebviewName type (exported from the same file) ensures compile-time validation of webview names.

// src/webviews/_integration/WebviewRegistry.ts
import { MyView } from '../../documentdb/myView/MyView';

export const WebviewRegistry = {
  collectionView: CollectionView,
  documentView: DocumentView,
  myViewName: MyView, // <-- add your entry
} as const;

export type WebviewName = keyof typeof WebviewRegistry;

Telemetry: publicProcedure vs publicProcedureWithTelemetry

BaseWhen to usectx.actionContext
publicProcedureFire-and-forget, no external calls, telemetry reported separatelyabsent (do not read it)
publicProcedureWithTelemetryDefault choice. Any procedure touching DB, network, or user-visible workGuaranteed, injected by the DocumentDB TelemetryRunner

publicProcedureWithTelemetry is publicProcedure.use(telemetryMiddlewareBody(documentDbTelemetryRunner, ...)) (built in trpc.ts). The framework's telemetryMiddlewareBody delegates to the DocumentDB TelemetryRunner, which wraps the call in callWithTelemetryAndErrorHandling, contributes the full IActionContext to ctx.actionContext, auto-generates a telemetry event named documentDB.rpc.{type}.{path}, and records errors, duration, and abort status.

actionContext is not a field on the base RouterContext — it is an additive enrichment. Narrow to WithTelemetry<RouterContext> (= RouterContext & { actionContext }) in an instrumented procedure to read it; a plain publicProcedure procedure narrows to bare RouterContext, so reading actionContext there is a compile error instead of a runtime undefined.

Access telemetry safely:

const myCtx = ctx as WithTelemetry<RouterContext>;
myCtx.actionContext.telemetry.properties.myCustomProp = 'value';
myCtx.actionContext.telemetry.measurements.itemCount = items.length;

AbortSignal Support

Every tRPC operation (query, mutation, subscription) receives its own AbortController. Cancellation flows:

Client (React)                              Server (Extension Host)
──────────────                              ──────────────────────
// Queries/Mutations:
ac = new AbortController()
trpcClient.myProc.query(input,
  { signal: ac.signal })
ac.abort()  →  sends 'abort' msg  →  abortController.abort()
                                        → ctx.signal.aborted = true

// Subscriptions:
sub = trpcClient.mySub.subscribe(...)
sub.unsubscribe()  →  'subscription.stop'  →  abortController.abort()

Using abort in procedures

// Pass signal to APIs that accept it (MongoDB driver, fetch, etc.)
getData: publicProcedureWithTelemetry
    .input(z.object({ filter: z.record(z.unknown()) }))
    .query(async ({ input, ctx }) => {
        const myCtx = ctx as RouterContext;

        // Option 1: Pass to driver (preferred)
        const cursor = collection.find(input.filter, { signal: myCtx.signal });

        // Option 2: Manual check in loops
        for (const item of items) {
            if (myCtx.signal?.aborted) return;
            await processItem(item);
        }
    }),

Client-side abort

const trpcClient = useTrpcClient();
const abortControllerRef = useRef<AbortController>();

const runQuery = async () => {
  abortControllerRef.current?.abort(); // cancel previous
  const ac = new AbortController();
  abortControllerRef.current = ac;

  const result = await trpcClient.mongoClusters.collectionView.myQuery.query(input, { signal: ac.signal });
};

When publicProcedureWithTelemetry detects an aborted signal, the DocumentDB TelemetryRunner sets telemetry.properties.aborted = 'true' and result = 'Canceled' automatically.

Subscriptions

Subscriptions stream multiple values from server to client using async generators:

// Server (router)
streamData: publicProcedureWithTelemetry
    .input(z.object({ batchSize: z.number() }))
    .subscription(async function* ({ input, ctx }) {
        const myCtx = ctx as RouterContext;

        for (let i = 0; i < total; i += input.batchSize) {
            if (myCtx.signal?.aborted) return; // check before each yield
            const batch = await fetchBatch(i, input.batchSize);
            yield batch;
        }
    }),

// Client (React)
const sub = trpcClient.mongoClusters.myView.streamData.subscribe(
    { batchSize: 100 },
    {
        onData(batch) { /* handle each batch */ },
        onComplete() { /* all done */ },
        onError(err) { /* handle error */ },
    },
);

// To stop:
sub.unsubscribe();

Client-Side Hook Usage

import { useTrpcClient } from '../_integration/useTrpcClient';
import { useConfiguration } from '@microsoft/vscode-ext-webview/react';

export const MyComponent = () => {
  const trpcClient = useTrpcClient();
  const config = useConfiguration<MyViewConfig>();

  useEffect(() => {
    trpcClient.mongoClusters.myView.getData.query({ id: config.documentId }).then(setData);
  }, []);
};

useConfiguration<T>() retrieves the initial config passed to WebviewController constructor (serialized via encodeURIComponent(JSON.stringify(...))).

Common Pitfalls

  • Never use any in procedure context casts — narrow with ctx as WithTelemetry<RouterContext> when the procedure reads telemetry (ctx.actionContext.telemetry), or ctx as RouterContext otherwise
  • Always prefer publicProcedureWithTelemetry unless you have a specific reason not to
  • Always check myCtx.signal?.aborted in long-running loops — not checking causes wasted work after client cancels
  • Do not mutate the shared context object — WebviewController clones it per-operation already, but router code should treat ctx as read-only
  • Input validation uses zod — always define .input(z.object({...})) for type safety
  • The commonRouter handles cross-cutting concerns (error reporting, telemetry events, surveys, URL opening) — do not duplicate these in view-specific routers

Plus de skills de microsoft

oss-growth
microsoft
Persona de growth hacker OSS
agent-framework-azure-ai-py
microsoft
Créez des agents Azure AI Foundry à l’aide du SDK Python Microsoft Agent Framework (agent-framework-azure-ai). À utiliser lors de la création d’agents persistants avec AzureAIAgentsProvider, de l’utilisation d’outils hébergés (interpréteur de code, recherche de fichiers, recherche web), de l’intégration de serveurs MCP, de la gestion de fils de conversation ou de l’implémentation de réponses en streaming. Couvre les outils de fonction, les sorties structurées et les agents multi-outils.
development
airunway-aks-setup
microsoft
Configurez AI Runway sur AKS — du cluster nu au modèle en cours d'exécution. Couvre la vérification du cluster, l'installation du contrôleur, l'évaluation GPU, la configuration du fournisseur et le premier déploiement. QUAND : « configurer AI Runway », « intégrer un cluster AKS », « installer AI Runway », « configuration airunway », « déployer un modèle sur AKS », « inférence GPU sur AKS », « configuration KAITO sur AKS », « exécuter LLM sur AKS », « vLLM sur AKS », « configurer le service de modèles sur AKS », « contrôleur AI Runway ».
devops
appinsights-instrumentation
microsoft
Conseils pour instrumenter les applications web avec Azure Application Insights. Fournit des modèles de télémétrie, la configuration du SDK et des références de configuration. QUAND : comment instrumenter une application, SDK App Insights, modèles de télémétrie, qu'est-ce qu'App Insights, conseils sur Application Insights, exemples d'instrumentation, bonnes pratiques APM.
devops
applicationinsights-web-ts
microsoft
Instrumentez les applications navigateur/web avec le SDK JavaScript Application Insights (@microsoft/applicationinsights-web). Utilisez-le pour la surveillance des utilisateurs réels (RUM) — vues de page, clics, dépendances AJAX/fetch, exceptions, événements personnalisés et traces d’agents GenAI côté navigateur corrélées aux traces OpenTelemetry backend. Couvre le script de chargement du SDK et la configuration npm, les extensions de framework (React, React Native, Angular), Click Analytics, les initialiseurs de télémétrie et les conventions sémantiques OTel GenAI pour les spans d’agents/outils/modèles émises depuis le navigateur.
devops
azure-ai-anomalydetector-java
microsoft
Créez des applications de détection d'anomalies avec le SDK Azure AI Anomaly Detector pour Java. Utilisez-le lors de l'implémentation de la détection d'anomalies univariées/multivariées, de l'analyse de séries temporelles ou de la surveillance basée sur l'IA.
development
azure-ai-language-conversations-py
microsoft
Implémentez la compréhension du langage conversationnel (CLU) à l’aide du SDK Python azure-ai-language-conversations. Utilisez-le lorsque vous travaillez avec ConversationAnalysisClient pour analyser l’intention et les entités d’une conversation, créer des fonctionnalités de NLP ou intégrer la compréhension du langage dans des applications.
development
azure-ai-ml-py
microsoft
SDK v2 d’Azure Machine Learning pour Python. Utiliser pour les espaces de travail ML, les tâches, les modèles, les jeux de données, le calcul et les pipelines. Déclencheurs : « azure-ai-ml », « MLClient », « espace de travail », « registre de modèles », « tâches d’entraînement », « jeux de données ».
development