What this tests
parsing
Point your client at this specimen and find out — before prod does — whether it actually validates what arrives, or just hopes.
Anatomy
The body is Latin-1. The header swears it's UTF-8. Every é, ñ, and ü arrives as a mojibake piñata — "café" rendered "café", which is either a corrupted string or a very small insurance company. Classic cause: a pipeline hop that transcoded the bytes but forgot to update the label — the cargo was repackaged and nobody changed the manifest. The bytes are self-consistent, the metadata is lying, and your decoder trusts the metadata, so out comes the Ã's like a ransom note assembled by a very tired linguist. Clients that sniff or attempt a fallback decode survive. Clients that trust Content-Type implicitly will faithfully store "Señor Data" in your CRM, where it will remain until a customer writes in asking why their name looks like it was translated into cat.
Replay it
curl -i https://faultydata.com/r/FD-2026-014
// response appears hereSignature
{Status:200 Headers:map[Content-Type:application/json; charset=utf-8] Body: Corrupt:encoding Delay:<nil> Flaky:<nil> Sequence:<nil> Drip:<nil> StallAfter:0 Hang:false Reset:false Raw: Phases:[]}
Keep digging
FD-2026-011 — Truncated JSON body FD-2026-012 — Malformed JSON (trailing comma) FD-2026-013 — Wrong types in a valid envelope FD-2026-015 — UTF-8 BOM prefix FD-2026-028 — Malformed channel file update (CrowdStrike, Jul 2024) FD-2026-034 — HTTP 200 wrapping an error (GraphQL/SOAP pattern)