Add gnaw: a Rust stack language with built-in CTF byte transforms

[warlyjr-cloud]

Hi! Adding gnaw — I think it fits the same slot as Kitten and Stabel on the list: it borrows Forth's model without claiming to be one.

What it is: a small concatenative language in Rust. Whitespace-separated words, a data stack, : name ... ; definitions, a separate return stack (>r r> r@), if/else/then, begin/until, counted do/loop with i, and a see disassembler that dumps a word's compiled bytecode. Lexer → single-pass compiler (jumps resolved with a stack of pending back-patches, no AST) → bytecode VM with a program counter.

What it is not: ANS-compliant. No IMMEDIATE/POSTPONE, no CREATE/DOES>, no vocabularies. I've tried to say that plainly in the README so nobody arrives expecting a standards Forth.

Its actual angle: built-in words for the byte transforms that come up constantly in CTF and reverse-engineering work — hex>bytes/bytes>hex, xorkey (repeating key), rot13, b64encode/b64decode, and brute (all 256 single-byte XOR keys, filtered by a natural-language heuristic). They all share an ( addr len -- addr len ) shape, so they compose, and slurp reads a real file into that same shape. Decoding a challenge ends up as one line:

"051807040f010d15113c1d153c150a0a02031e" hex>bytes "ctf" xorkey type cr
→ flag{gnaw_is_alive}

There's a browser playground (WebAssembly, runs locally) if you'd rather poke at it than clone it.

Totally understand if it's out of scope for the list — thanks either way for maintaining it.