Hello,

I’m the team owner on a Hobby plan and need clarification about deployment-file recovery.

A single authenticated request to the deployment file-content API returned valid JSON/base64. The decoded file is 245,923 bytes, including its final newline, and its SHA-1 matches the requested file UID.

Earlier dashboard exports were 245,922 bytes. A read-only diagnostic excluding the API file’s final newline reproduces both recorded hashes of those earlier exports.

The Vercel AI support agent initially stated 38,971 bytes, then retracted that figure as a truncated-preview calculation. It offered a prefilled engineering case, but the case form blocks submission under CI/CD → Builds on Hobby.

Could Vercel staff clarify:

1. Does the deployment file-content endpoint return the originally uploaded source bytes or potentially transformed build output?
2. Can the dashboard source export omit the final newline?
3. If this requires inspection of the specific deployment, what supported private route is available for providing its identifiers?

The original files and request receipts are preserved. I can provide identifiers privately through a staff-approved channel.

Thank you, Wesley