feat(ci): promote Trigger.dev tasks in lockstep with the ECS traffic cutover

[TheodoreSpeaks]

Summary

• Move Trigger.dev staging + production deploys into CI (off the git-integration auto-deploy), deploying with --skip-promotion so new task runs keep executing the old version until we promote.
• Promote the deployed version at the exact ECS traffic cutover: a new promote-trigger-* job polls CodeDeploy (AllowTraffic) and then runs trigger promote, so tasks and app cut over in lockstep instead of tasks flipping ~4–50 min early (the prod bake queue).
• Correlate this push to its ECS deploy via the app image digest (build-amd64 publishes it as an artifact) → CodePipeline execution → CodeDeploy deployment id → AllowTraffic. ECR deploy tags are floating, so the digest is the only durable key.
• Gate promotion on promote-images.promoted: when the stale-run guard skips the image retag, tasks correctly stay on the old version. If the app deploy fails or never cuts over, promote never fires and the job fails visibly.
• New .github/scripts/wait-for-ecs-cutover.sh does the poll; verified end-to-end against real staging deploys (digest → d-… → AllowTraffic).

Prerequisites before this runs on staging/prod

• Add CodePipeline + CodeDeploy read perms to the OIDC deploy roles (STAGING_AWS_ROLE_TO_ASSUME, AWS_ROLE_TO_ASSUME), which are managed outside the repo: codepipeline:ListPipelineExecutions/GetPipelineExecution/ListActionExecutions, codedeploy:GetDeployment/ListDeploymentTargets/GetDeploymentTarget.
• Set the TRIGGER_ACCESS_TOKEN repo secret (a Trigger.dev PAT; spans all environments). Then disable the Trigger.dev git integration for staging/prod.

Type of Change

• New feature (CI/CD)

Testing

• Ran wait-for-ecs-cutover.sh against real completed staging deployments: correctly resolved digest → pipeline execution → CodeDeploy deployment → AllowTraffic (exit 0), and timed out cleanly on an unmatched digest (exit 1).
• actionlint clean (only pre-existing custom-runner-label warnings); bun run lint and bun run check:api-validation:strict pass; YAML validated.

Checklist

• Code follows project style guidelines
• Self-reviewed my changes
• Tests added/updated and passing
• No new warnings introduced
• I confirm that I have read and agree to the terms outlined in the Contributor License Agreement (CLA)

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated (UTC)
docs	Ready	Preview, Comment	Jul 19, 2026 9:00pm

[cursor]

PR Summary

Medium Risk
Changes production deploy sequencing for app and background tasks; mis-correlation or missing AWS read IAM could block promotion or delay cutover, but app deploy path is unchanged and stale-run guards limit accidental early promote.

Overview
Moves main/staging Trigger.dev deploys into CI and delays task promotion until the ECS app has finished its blue/green traffic shift, so background tasks and the app cut over together instead of tasks flipping minutes earlier.

A new deploy-trigger job uploads task versions with --skip-promotion (parallel to image build, not blocking app deploy). build-amd64 publishes the app image digest as an artifact; promote-images now exports promoted and retag_epoch when deploy tags move. promote-trigger runs only when promotion actually happened, polls via .github/scripts/wait-for-ecs-cutover.sh (digest + retag time → CodePipeline → CodeDeploy AllowTraffic on all targets), then runs trigger promote for the version from deploy-trigger. If ECS never cuts over or the stale-run guard skips retagging, promotion is skipped or the job fails visibly.

^{Reviewed by Cursor Bugbot for commit f7135b0. Bugbot is set up for automated code reviews on this repo. Configure here.}

[greptile-apps]

Greptile Summary

This PR coordinates Trigger.dev task promotion with the ECS traffic cutover. The main changes are:

• Deploy Trigger.dev versions from CI without immediately promoting them.
• Publish the app image digest and correlate it with the matching CodePipeline execution.
• Wait for successful AllowTraffic events across all ECS deployment targets.
• Promote the Trigger.dev version only after the app traffic shift.

Confidence Score: 5/5

The latest fixes look safe to merge.

• The timestamp filter prevents old same-digest deployments from satisfying the normal cutover gate.
• The cutover loop now requires successful AllowTraffic status for every returned ECS target.
• No additional blocking issue met the scope for this follow-up review.

Important Files Changed

Filename	Overview
.github/scripts/wait-for-ecs-cutover.sh	Adds polling that resolves the matching CodeDeploy deployment and waits for every ECS target to complete its traffic shift.
.github/workflows/ci.yml	Adds staged Trigger.dev deployment and promotion jobs tied to the ECS image promotion workflow.

_{Reviews (3): Last reviewed commit: "fix(ci): widen promote-trigger job timeo..." | Re-trigger Greptile}

[TheodoreSpeaks]

@greptile review

[TheodoreSpeaks]

@cursor review

[cursor]

AWS creds expire before poll

High Severity

promote-trigger assumes the deploy role with the default ~60-minute session and never sets role-duration-seconds, while wait-for-ecs-cutover.sh is built to poll for up to 70 minutes (the prod bake-queue case). After the session expires, AWS calls fail; those errors are swallowed, so the script keeps retrying until timeout and never promotes. The long-wait path this change was written for leaves the app cut over while Trigger.dev stays on the old version.

Additional Locations (1)

• .github/scripts/wait-for-ecs-cutover.sh#L26-L27

 

^{Reviewed by Cursor Bugbot for commit f7135b0. Configure here.}

[TheodoreSpeaks]

@greptile review

[TheodoreSpeaks]

@cursor review

[cursor]

Cursor Bugbot has reviewed your changes using default effort and found 1 potential issue.

There are 2 total unresolved issues (including 1 from previous review).

^{❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, enable autofix in the Cursor dashboard.}

^{Reviewed by Cursor Bugbot for commit f7135b0. Configure here.}

[cursor]

Same digest blocks task promotion

Medium Severity

Phase A requires a CodePipeline execution for the app digest with startTime at/after retag_epoch. That rejects the historical same-digest run this filter was added for, but retagging latest/staging to a digest it already points at is often a no-op for ECR (no new EventBridge/pipeline). The poll then never finds a post-retag execution and times out, so the skip-promoted Trigger.dev version from deploy-trigger is never promoted—especially on re-runs after promote-images already moved the tag, or any push that reuses the app image digest (Trigger tasks live in background/ and are not in the final app image).

Additional Locations (1)

• .github/workflows/ci.yml#L514-L522

 

^{Reviewed by Cursor Bugbot for commit f7135b0. Configure here.}