We need read-only assistance with a production environment before deploying an already-reviewed code update.

During our metadata-only investigation:

• The service was running a successful deployment.
• GitHub autodeploy was connected to main, with Wait for CI disabled.
• environmentStagedChanges returned an EnvironmentPatch with status STAGED.
• The corresponding project/environment canvas showed no staged-changes banner, Details button, or visible change markers.
• environment.unmergedChangesCount returned null.
• Connector summaries reported changeCount=47 at environment level and 44 at service level, but their counting semantics were not documented.
• The service projection returned staged.config={} and a list of variable names. We do not consider this proof of either an empty or nonempty effective diff.

No configuration changes, apply/discard operations, deployments, restarts, or rollbacks were performed during this investigation.

Could a Railway team member clarify:

1. Can this STAGED object exist when there are no effective pending changes? How can we verify that for our environment?

2. Would the next GitHub-triggered deployment after a push to main consume this uncommitted STAGED patch, or use the last committed configuration? Please distinguish GitHub autodeploy from “Deploy staged changes” and redeploying an existing deployment.

3. Is there a supported read-only method to obtain the effective diff as field paths and add/update/remove operations, without returning variable values? If not, could your team verify the effective diff for us?

We can provide the exact project, environment, service, patch, and deployment IDs through a private support channel. Please do not make any changes to the project.

Generic instructions to open “Details” do not resolve this case because the staged-changes banner is absent.