[jayleaton]

Bug

README "Backups and restore" says Railway's volume snapshot schedule "is configured in the dashboard only (the CLI and IaC cannot set it yet)". That's wrong: Railway's public GraphQL API sets it, and a derived project sets the schedule through it on every apply.

Reproduction

A read-only introspection of Railway's public API, with no auth and no project touched:

$ curl -s https://backboard.railway.com/graphql/v2 -H 'content-type: application/json' \
    -d '{"query":"{ __type(name:\"Mutation\") { fields { name args { name } } } }"}'  # filtered to *BackupSchedule*
volumeInstanceBackupScheduleUpdate ['kinds', 'volumeInstanceId']
$ ... '{ __type(name:"VolumeInstanceBackupScheduleKind") { enumValues { name } } }'
{"data":{"__type":{"enumValues":[{"name":"DAILY"},{"name":"MONTHLY"},{"name":"WEEKLY"}]}}}

Fix

Doc only. The sentence now says Railway setup doesn't configure the schedule, and that you set it in the dashboard (the database's Backups tab) or with volumeInstanceBackupScheduleUpdate(volumeInstanceId, kinds: [DAILY, WEEKLY, MONTHLY]). The IaC file still can't express it. Managing the schedule from platform.yaml would be a feature, so it isn't in this PR.

Verified

• make check-fast: ok (docs-only, 0 packages).
• No other doc repeats the claim (grep -rn "dashboard only|cannot set it" over README, docs/, AGENTS.md, CONTEXT.md, deploy/).

🤖 Generated with Claude Code

[jayleaton]

Superseded by #122 (merged as e80a772). The README section moved to docs/backups.md in #107, and #122 rewrote it: snapshot schedules are now declared in deploy/platform.yaml and set through Railway's public API (volumeInstanceBackupScheduleUpdate), with the dashboard as the fallback. Closing at JJ's request.