COBS decode reads the space or comma separated hex bytes encode-cobs prints, then walks the length-prefixed blocks to strip the framing and restore any zero bytes COBS stuffed out. A zero byte where a length code belongs, or a block running past the input's end, is refused with its offset. It expects that hex-byte format, never raw text.
Questions
What input does it expect?
Space or comma separated hex bytes, the exact text format encode-cobs prints, not raw text.
What happens with a zero byte where a length code belongs?
The tool refuses with "unexpected zero byte in COBS stream" and the offset.
What happens with a truncated block?
The tool refuses, naming how many more bytes the block needed.
What is COBS for?
Consistent Overhead Byte Stuffing removes zero bytes from a stream so a protocol can use zero as a frame delimiter.
Can I use this from the command line?
Yes, the same code runs on the page, the CLI and MCP.
Related Dev tools
n→nBase64 encodeBase64 encode.Runs on each file you dropn→nBase64 decodeBase64 decode.Runs on each file you dropn→nBase64URL encodeBase64URL encode.Runs on each file you dropn→nBase64URL decodeBase64URL decode.Runs on each file you dropn→nBase32 encodeBase32 encode.Runs on each file you dropn→nBase32 decodeBase32 decode.Runs on each file you dropn→nBase45 encodeBase45 encode.Runs on each file you dropn→nBase45 decodeBase45 decode.Runs on each file you drop