Beamer check: add initiate-l1-invalidations command

[istankovic]

This PR is part of #1646, adding the beamer check initiate-l1-invalidations command.

[istankovic]

Still missing; docs. 🤦

[fredo]

All in all, looks pretty good. I'd like to be aligned on my comment above though.

What's also missing (but part of the description in #1646) are the prove() calls for OP Stack chains.

[istankovic]

What's also missing (but part of the description in #1646) are the prove() calls for OP Stack chains.

Yeah, I realized that just after marking this ready for review. 🤦
Will add, of course. (I'm not sure how to even approach testing this part so I'm probably going to leave it out of this PR. Maybe something based on top of our existing OP E2E tests, but even that does not seem fully straightforward...)

[fredo]

What's also missing (but part of the description in #1646) are the prove() calls for OP Stack chains.

Yeah, I realized that just after marking this ready for review. 🤦 Will add, of course. (I'm not sure how to even approach testing this part so I'm probably going to leave it out of this PR. Maybe something based on top of our existing OP E2E tests, but even that does not seem fully straightforward...)

I would be fine to push this into another PR cause we need to think a bit more about the information being passed here. For OP chains the finality period only starts after the prove() call was made, so we cannot really add the timestamp if the prove() didnt happen.

Alos here another concern is, that i.e. in the agent we try to call the prove() regardless of the state root being published and simply retry. But this also takes a couple of minutes. So we may want to even think about pushing the prove out of the invalidation creation?
Otherwise I can see that after every invalidation the script will halt until the state root has been published. We may want to consider flagging the invalidations which need to be proven and do this in a separate step or we need to properly handle the waiting time and the relayer exiting with an exception (if the state root is not published yet)

With regards to testing, in order to really test it we would need an actual OP setup like in the e2e tests...

[istankovic]

Alright, the docs are now also ready. 📜

[fredo]

very nice 🍷