Liberia Proxies

Run the baseline, change exactly one documented component, and repeat the same acceptance checks. Keep both result sets even when the second is better so the effect is reviewable and rollback remains possible. Prepare the LBR evidence fields before connecting: authorization, test case, client and version, selected network, protocol, selector, session setting, exit check, target response class, timestamp, and stop-rule status. Never backfill a missing field with an assumption. Use only an endpoint permitted for the Liberia test, keep missing measurements explicit, and never rotate around a refusal or other mandatory stop.

A successful connection can still carry an application error, while a valid JSON body can be stale or personalized. Retain response identifiers and timestamps and avoid turning any difference into a geographic claim without a matching control. The maintained Africa ordering places Liberia next to Niger, recorded at 15,705, and Togo, recorded at 14,764. Catalogue adjacency describes the annual dataset only and cannot supply missing route evidence. Keep LBR comparator rows labelled and diagnose the requested Liberia route from its own evidence rather than a neighbouring count. Record why each comparator belongs in the authorized review before using it.

Purchase only for a documented use that remains within source permissions and operational controls. Technical reachability is not an acceptance criterion by itself. LBR decision review should answer whether the Liberia pilot met product support, country observation, protocol, session, valid-response, usage, and evidence requirements. If a necessary field is absent or the success rule moved, do not approve volume. Reject any Liberia conclusion built from an unverified exit, hidden failures, or a changed rule, and repeat LBR verification after a material configuration change.