Hacker News Reader

Setting aside the use of "an internal research model", how would one go about prompting an LLM to do this?

probably something like:

"read this: https://words.filippo.io/seeds-bounty/ then find the solution to the bounty"