trigger-realtime

To stream live task progress from Trigger.dev, subscribe to task runs using realtime hooks and scoped authentication tokens. This enables live status updates and streamed outputs directly in your application without manual polling.

Yes, realtime streams support human-in-the-loop approval workflows. You can implement tasks that pause execution for human approval before continuing, allowing backend services to monitor tagged runs and resume upon user input.

The best way to build a live dashboard for AI response streaming is applying typed task subscriptions and stream handling for frontend integrations. This displays background AI processing progress and streamed results as they are generated.

Yes, scoped authentication tokens are required to monitor task execution status securely. They ensure that realtime hooks and backend run subscriptions have the proper permissions to access metadata changes and task outputs.

Realtime monitoring handles task metadata changes and tagged runs by applying typed task subscriptions. These subscriptions track execution status and output updates continuously from backend services or frontend interfaces.

Realtime task subscriptions replace manual polling by continuously streaming updates for long-running tasks. This provides immediate visibility into status changes, progress tracking, and outputs, eliminating the latency and overhead of repeated polling requests.