Publish Events with Azure Event Grid SDK for Python

Wraps azure-eventgrid for Python to publish events for event-driven architectures using Azure Event Grid, an event-routing service with pub/sub semantics. It authenticates an EventGridPublisherClient via DefaultAzureCredential against a topic endpoint (EVENTGRID_TOPIC_ENDPOINT) or, for Event Grid Namespaces, a namespace endpoint (EVENTGRID_NAMESPACE_ENDPOINT).

event = CloudEvent(
    type="MyApp.Events.OrderCreated",
    source="/myapp/orders",
    data={"order_id": "12345", "amount": 99.99}
)
client.send(event)

It supports two event formats: CloudEvents 1.0 via the CloudEvent class (the recommended, standard, interoperable format, with required type/source and optional subject, datacontenttype, dataschema, time, and custom extensions), and Azure's native Event Grid Schema via the EventGridEvent class (required subject, event_type, data, data_version, with an auto-set topic and optional event_time). Publishing covers a single event or a batch - a list of events sent in one client.send() call. A fully async client (azure.eventgrid.aio) mirrors the sync API with await and an async context manager. For Event Grid Namespaces (pull-delivery topics), the client connects to the namespace endpoint instead of a topic endpoint and passes the target namespace_topic name on each send.

Stated best practices: use CloudEvents for new applications since it's the industry standard, batch events when publishing multiple at once, include meaningful subjects for downstream filtering, use the async client for high-throughput scenarios, rely on Event Grid's built-in retry rather than hand-rolling one, and set event types deliberately since they drive routing and filtering on the subscriber side.

Use it for publishing events, handling CloudEvents, or building event-driven architectures on Azure Event Grid from Python. As with any skill output, verify results against the current environment, and stop to ask for clarification if required inputs, permissions, safety boundaries, or success criteria are missing.

Input is an Event Grid topic or namespace endpoint/credential and one or more events (CloudEvent or EventGridEvent) with their type, source or subject, and payload - plus optional CloudEvent fields like datacontenttype, dataschema, time, and custom extensions, or the EventGridEvent's required data_version and auto-set topic. Output is the event(s) published to the topic, ready for Event Grid to route to subscribers.

Uses the azure-eventgrid SDK, in both its sync and async (azure.eventgrid.aio) forms, and azure-identity's DefaultAzureCredential (also available as an async credential) for authentication, publishing in either the CloudEvents 1.0 standard or Azure's native Event Grid Schema, to either custom topics or Event Grid Namespaces.

Python developers building event-driven applications on Azure who need to publish CloudEvents or Event Grid events reliably, individually or in batches.