I highly recommend pi.dev for doing anything creative with agents and harnesses. It separates the harness from the model, so you need to BYOM keys and subscriptions.

You can build your own harness quite easily or use a harness something like pi.dev which offers a lot of flexibility in customization in terms of tools, models ised, resume sessions at a prior point in time, etc.

I have used pi.dev to easily start from a known context and goal and use it to compare the cost effectiveness, token isage, etc. of different models to reach the goal state.

I am sure there are many ways to do this successfully.

It's not clear whether the author used Google's resources or his own personal account to run the queries.

Type into claude code:

"Scan through publically accessible LIDAR data on <region of mexican jungle>, search for potential Mayan ruins and cross reference with published data to rule out existing finds

Work autonomously until the entire area is searched, prepare me a report detailing potential discoveries"

should more or less do the trick