Issue

All deployments are being aborted at the platform level immediately after container creation, before the build ever starts. No build logs exist and the deployment pipeline never advances past CREATE_CONTAINER.

Technical Details

• Timeline: Every deployment since the last successful build (commit bd4e517, 16:55 UTC) has failed with the same signature — approximately 8 consecutive failed deployments over the past hour+
• Pipeline Stage: Only CREATE_CONTAINER completes; pipeline never advances to SNAPSHOT_CODE or BUILD_IMAGE
• Build Logs: Completely empty — no error output, no build stages recorded
• Error Message: Generic "Express deployment failed" with null failureStage, failureError, and failedReason; failure stage marked as "manual," indicating a platform-level abort
• Build Config: Node 22, Nuxt, npm (identical to last successful deployment; no config drift)
• Config Files: No railway.toml, railpack.json, or Dockerfile in repo — using Railpack defaults
• Prior Testing: Redeployed the exact previously-successful source tree (commit bd4e517) and it also failed identically, ruling out code or config regression

What Was Already Tried

• Recent commits attempted to fix a suspected Nuxt build OOM (4GB Node heap, disabled sourcemaps)
• These changes are chasing a failure that never reaches the build stage — real build OOM would produce BUILD_IMAGE logs with kill signal; none exist
• Redeploy of the last known-good commit also failed identically