Trigger.dev v4.5.12, released on August 21, 2026, introduces a key new feature: configurable execution windows for declarative scheduled tasks, alongside 5 improvements, 4 bug fixes, and 17 server changes to enhance task reliability and developer experience, according to Trigger.dev's changelog.
This update focuses on refined control over background jobs and a more robust platform. The new execution window feature provides developers with greater flexibility in managing task timing, moving beyond strict CRON schedules.
The release also brings significant enhancements to deployment workflows and overall system stability, streamlining operations for both cloud and self-hosted users.
What are the key updates in Trigger.dev v4.5.12?
Trigger.dev v4.5.12 primarily introduces configurable execution windows for declarative scheduled tasks, a new feature allowing tasks to run within a defined time slot rather than at a precise CRON time. This ensures stable execution offsets across invocations. The release also includes 5 improvements, 4 critical bug fixes, and 17 server-side enhancements, bolstering the platform's reliability and developer tooling.The execution windows are designed to assign a stable slot to each task run, maintaining consistency. Both the nominal CRON trigger time and the assigned execution time are exposed via the Schedule API, offering clear visibility into task scheduling.
The dashboard now visually represents these configured windows, along with upcoming assignment times, simplifying task management for developers.
How do new deployment controls enhance development?
New deployment controls in Trigger.dev v4.5.12 significantly enhance development workflows by introducing external ID tagging and updated Node.js defaults. Developers can now tag deployments with identifiers like commit SHAs, streamlining tracing and version management. Additionally, new projects default to Node.js 24, while legacy projects using Node.js 21 can be easily identified for updates.The `trigger.dev deploy --external-id` command allows developers to assign a custom identifier to each deployment, such as a commit SHA or CI run ID. This feature aids in tracking specific releases and enables rebuilding with a `--force` flag when the same ID is deployed twice, as described by Trigger.dev.
Feature | Previous Behavior | v4.5.12 Enhancement |
|---|---|---|
Deployment Tagging | Limited version tracking | External ID tags (e.g., commit SHA) for deployments |
Node.js Default | Node.js 21 for new projects | Node.js 24 for new projects |
Runtime Pinning | Runs could execute on wrong versions | Runs pinned to specific deployment version |
For version protection, `TRIGGER_EXTERNAL_DEPLOYMENT_ID` pins runs to the originating deployment. On platforms like Vercel, automatic skew protection is enabled with `TRIGGER_AUTOMATIC_SKEW_VERSION_PROTECTION=1`, detecting commits to prevent an old release from triggering tasks from a newer one. This improves workflow integrity, which is crucial for modern CI/CD pipelines, much like the unified AI spend tracking offered by Vercel's new API.
How does Trigger.dev boost reliability and performance?
Trigger.dev v4.5.12 significantly boosts reliability and performance through 4 bug fixes and 17 server changes addressing critical issues. These updates include resolving problems with missing task metrics, fixing idempotency key handling for 64-character keys, and preventing fair queue tenants from getting stuck. Server-side improvements enhance task triggering resilience and speed up dashboard operations.One notable fix prevents unrelated runs from merging into single traces in external observability tools, ensuring clearer debugging. Another bug fix resolves issues where task metrics would go missing for projects with custom metric exporters.
The `idempotencyKeys.reset()` function now correctly handles keys that are precisely 64 characters long, a previous point of failure. Fair queue tenants are also more resilient; leaked concurrency slots are now recovered, preventing messages from getting stuck or being run twice.
Server changes improve global log search speed and make task triggering more resilient to brief service interruptions. Failed AI SDK tool calls and embedding spans now display detailed error messages and stack traces in the run inspector, aiding developers in troubleshooting.
