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 JSON is structurally valid except for the NUL bytes scattered through the strings — \x00, the terminator, sitting in the middle of a field value like a landmine in a picnic. Strict JSON parsers reject it outright (control characters must be escaped), lenient ones read it and produce strings with invisible holes. Where it comes from: C-language backends that read fixed-width buffers and serialize the padding; file formats that store strings NUL-terminated; a pipeline hop that treated the payload as a C string, which it was, in the sense that everything is a C string if you're careless enough. The bytes are data to one layer and end-of-string to another — the same payload meaning different things at different layers, which is the entire history of injection bugs in one sentence. Clients that pass the body through C-adjacent code will see the string end early — "ad" instead of "ada" — and if that string is an ID, a name, or a checksum, the missing bytes are now your incident.
Replay it
curl -i https://faultydata.com/r/FD-2026-074
// response appears hereSignature
{Status:200 Headers:map[] Body: Corrupt:nulls 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-2026-014 — Encoding mismatch (claims UTF-8, sends Latin-1) FD-2026-015 — UTF-8 BOM prefix FD-2026-028 — Malformed channel file update (CrowdStrike, Jul 2024)