CBOR Decoder (Diagnostic Notation)

It's the human-readable text form of a CBOR data item defined in RFC 8949 §8 — similar to JSON but able to express everything CBOR can: byte strings as h'…', distinct integer vs float (1 vs 1.0), tags as number(content), the simple values undefined and simple(n), and indefinite-length items with an underscore, e.g. [_ 1, 2]. It's the standard way to show what a CBOR blob contains.

Anything that emits CBOR gives you bytes: a WebAuthn attestationObject, a COSE_Key or COSE_Sign1, an IoT sensor frame, etc. Capture those bytes and paste them as hex (with or without spaces or a 0x prefix) or base64. The tool ignores whitespace and accepts base64url.

CBOR distinguishes integers from floating-point numbers at the byte level, and diagnostic notation preserves that: the integer one is 1, while the floating-point one is 1.0. Half-precision floats (2 bytes), single (4), and double (8) all decode to their exact value, including NaN and ±Infinity.

A CBOR message is a single data item. If there are extra bytes after the first complete item, the tool reports them rather than silently ignoring them — usually it means the input is a sequence of items, was truncated mid-item, or the format (hex vs base64) is wrong.