x3CTF babyrng writeup

This is my writeup for the babyrng cryptography chall from x3CTF 2025. This was my first formal exposure to Haskell and x3CTF.

I didn’t do this live as I was engaged with some other work but dived right into it when I was free.

Hope you learn something new from this thorough documentation of my brain coping and seething with Haskell syntax and crypto math.

begin

The prompt:

A flag has been stolen, destroyed and buried underground in a random location. System.Random was used, so it should be impossible to recover the flag… right?

random and the chall name itself interested me and I assumed that it’d be probably about breaking some sort of pseudo-deterministic random bit generation (I prefer DRBG instead of PRNG, sounds cooler).

Downloaded the babyrng.tar.gz tarball, extracted it to find a Dockerfile, the source.hs, and output.txt. Didn’t have Haskell toolchain installed so started building the image. Ran a container with the image and got the output

The shredded flag (f942732a0041b935d69547486eafe8) has been buried at (17234233598314619506, 13831628640053652072)

Hmm, interesting — time to read the src now. source.hs

flag = "MVM{[REDACTED]}"
 
shred :: String -> State StdGen String
shred "" = return ""
shred (c:cs) = do
 
	k <- state $ uniformR (0, 255)
	((:) (chr $ (ord c) `xor` k)) <$> (shred cs)
 
burryTreasure :: State StdGen String
burryTreasure = do
	shredded <- shred flag
	x :: Word64 <- state uniform
	y :: Word64 <- state uniform
	return $ printf "The shredded flag (%s) has been buried at (%d, %d)" (shredded >>= (printf "%02x" :: Char -> String)) x y
 
main :: IO ()
main = do
	rng <- getStdGen
	putStrLn $ evalState burryTreasure rng

It was a ~20 line source file, from the looks of it not much going on we have a variable of type string defined which is the flag then some transformations on it which I can’t really make sense of.

What next? Deepseek? GPT?

Nah — we raw dog it. And so I went head first into Haskell docs and tutorials to try and make sense of the spaghetti I see.

OSINT it, we ball

After getting a feel for the Haskell syntax and ecosystem, wrote a few programs to get a better grasp on the syntax. Big shoutout to these videos 1,2,3, which gave me enough Haskell literacy to do the chall. Now with resume-mention-worthy syntax grasp, the actual chall reversing starts.

Clearly I’m not the sharpest Haskell engineer on the block so, we switch from cryptanalysis to OSINT mode. First, got an overview of the packages that are being used. Interesting ones:

Recent Writing