A detached volume (no service attached) cannot be deleted by any method - every path reports success but the volume remains listed with its instance alive. Project: som (f41225e8-503c-4eeb-9449-fbe3b7827bc4) - environment: production (f54ee520-3ffe-45a5-8666-ac360517ae9c) - volume: postgres-volume-R-MM (id 5e791b45-242a-47fa-a07a-0fd2093e4e54), ~199 MB, serviceId null. It does not render on the project canvas nor in Settings, so there is no manual delete button; it only shows via CLI "railway volume list" and the GraphQL API. Tried, in order: (1) GraphQL volumeDelete while detached - returns true, no effect, twice; (2) dashboard Agent staged removal + Deploy + Destructive Changes dialog (typed the name) + Commit - shows "Deployed - 1 resource removed", volume still there; (3) CLI "railway volume delete" - prints deleted, still there; (4) attached it to a fresh throwaway service and ran volumeDelete while attached - true, no effect; (5) deleted that throwaway service - service died, volume survived the cascade and returned to detached; (6) environmentPatchCommitStaged - "No patch to apply". The "-R-MM" suffix suggests a leftover from a deleted environment (an orphaned record the deletion handlers skip silently). Identical symptom to the solved thread "Phantom volume visible in CLI but not in dashboard" (b70764de), which has no public resolution posted. Could you hard-delete this volume server-side? Please do NOT touch the volume "postgres-volume" attached to the Postgres service in the same project - that one is live production data. Thanks!