Angola Proxies

Validate each response against the expected shape before comparing values. Keep credentials, parameters, pagination, and request body fixed; classify missing fields, type changes, timeouts, and application errors separately from location-related differences. For AGO, retain one configuration record and one result record for every attempt. Include target permission, selector, network, protocol, connection reuse, observed country, response marker, error boundary, and UTC time. Stop rather than rotate around an explicit denial. Use only an endpoint permitted for the Angola test, keep missing measurements explicit, and never rotate around a refusal or other mandatory stop.

Do not infer city, ZIP, coordinates, carrier, ASN, language, or resident identity from country success. If a narrower field matters, confirm that the chosen product supports it and validate that field in a separate test. The Africa list places Seychelles's 10,185 and Zambia's 9,313 nearest to the Angola historical total. That placement helps audit the catalogue; it neither confirms current supply nor ranks buyer outcomes. Keep AGO comparator rows labelled and diagnose the requested Angola route from its own evidence rather than a neighbouring count. Record why each comparator belongs in the authorized review before using it.

Buy for cache-oriented testing only after the pilot yields interpretable headers and repeatable content digests. If the destination hides cache evidence or changes independently, record the limitation instead of raising request volume. Before committing to Angola, run the smallest representative workload and normalize offers by minimum purchase, metering basis, valid outcomes, failed outcomes, and required session behavior. The AGO decision must use observed pilot evidence, not the annual address total as a performance proxy. Reject any Angola conclusion built from an unverified exit, hidden failures, or a changed rule, and repeat AGO verification after a material configuration change.