Run‑Length Encoding for UTF‑8 bytes. Client-side only. No uploads.
Format: tokens like 12x41 meaning 12 repeats of byte 0x41 (ASCII A). Tokens are separated by whitespace.
Use this when you need a tiny compression primitive for receipts/logs (e.g., repeated padding, long runs of spaces/zeros). It will not compress natural language.
5 MiB output by default (to avoid foot-guns).