http://chals.swampctf.com:50207/
https://ctf.swampctf.com/files/68f09c26e413cc219c7711c6e945d8fc/MaybeHappyEndingGPT.zip?token=eyJ1c2VyX2lkIjoxOTEsInRlYW1faWQiOjEwOCwiZmlsZV9pZCI6Mzd9.Z-oTmg.qLXXXrIORiDeBtnveM7ITCvmXr0
- We find references of flag
- We see that there is a hint here, and that it is using eval()
- It is reading the LLM’s response and then evaluating it and returning the eval’s output as a response
