Since ~23:43 UTC on 2026-08-28, new deployments on groupsales-api (f36b1a40-0339-4bc1-bde4-9925c776a6dd) and groupsales-admin (dcaaf0c8-842c-4a79-8b32-9b62cd6e40dd) in project

cd4c71b0-a766-41ab-b4e3-e3445d9050da, both sandbox and production environments, have been stuck at INITIALIZING/BUILDING for 10+ minutes with no resolution. railway up uploads fine but

the deployment never gets a build (railway logs --build → "Deployment does not have an associated build"); the dashboard's redeploy also fails with "cannot re-deploy without a

snapshot." Switching a service to GitHub-triggered deploys didn't help either — one attempt briefly reached BUILDING before also stalling. Example deployment IDs:

e56bd710-f24c-4a7f-859f-701c9b47b2df, 1fd77adf-3bb9-4621-b4a5-efd0738a83aa, b26cb256-8b62-4ab0-926f-c173714b9d46.

Worse: reconnecting sandbox groupsales-api's GitHub source stopped its previously-running healthy instance without the replacement ever coming up — that service currently has no running

instance. Production is still on its last-good deployment and unaffected, but we're blocked from shipping a fix that's needed for a demo. Could you look into why deployments aren't

progressing past initialization, and why the source-reconnect dropped a running instance without failover?