Hello,
I am experiencing a persistent publication failure when deploying a full-stack Google AI Studio app to Cloud Run in europe-west2.
The application passes all checks inside the AI Studio development environment:
-
Dependency installation succeeds.
-
TypeScript validation passes.
-
Production build succeeds.
-
The server starts successfully.
-
Health endpoints return HTTP 200.
-
Authentication, email delivery, and core application functions work in the development environment.
However, every publication attempt fails during Cloud Run service creation with:
The provided source archive is corrupted.
Cloud Run reports that the base image is valid, but configuration readiness fails. The application never reaches npm installation, server startup, or its startup probe. This indicates that the AI Studio-generated publication archive is rejected before the application executes.
I have already:
-
Created a fresh AI Studio app.
-
Imported the application from a validated GitHub recovery repository.
-
Reinstalled and synchronized dependencies.
-
Reduced unnecessary dependencies and native binaries.
-
Verified the production build and development runtime.
-
Revalidated all required application assets.
The same corrupted-source-archive error still occurs on the first publication attempt of the fresh app.
This appears to be an AI Studio Build/Publish packaging or upload problem rather than an application-code failure.
Could the AI Studio team please confirm whether this is a known issue and advise how to force regeneration of the managed publication archive without recreating or modifying the validated application again?
I can provide the App ID, project identifiers, managed archive URI, revision, operation ID, sanitized audit log, and exact timestamp privately to a verified Google support representative.
Thank you.