Time-To-Live limits documented for background tasks

Background tasks are great until a queue backs up and jobs execute long after they are relevant. To prevent stale executions, the documentation now outlines how to set Time-To-Live (TTL) limits across three different levels: globally in the project config, directly on a task definition, or overridden on a specific trigger.

If a run sits in the queue longer than its specified TTL, it automatically expires. This is particularly useful for frequent scheduled jobs—if a long-running sync delays the next scheduled execution, the delayed run simply drops rather than stacking up. Developers can selectively opt tasks out of global TTL defaults by passing ttl: 0. These documentation changes serve as the companion guide to the newly released TTL packages.