Best practice developing for Mips on x86_64?

Hi folks,
I'm new to OpenWRT and need some guidance and help on how to be more efficient. I have a feeling I'm doing completely wrong...
Please, give me hints or suggestions on how to do what I describe below more efficiently.
For more than a month I've been digging into OpenWRT and a lot of home-made packages for a customized hardware built upon a PRX321 processor. There is no one left to ask about the code-base. I have a Docker-container with a build-environment and it works to build and run on target.
It feels like I've been thrown back to the 90s when inserting printf's, rebuilding the image for an hour, analyzing and doing changes in the packages and then rebuild again. So I decided to spend this week to get hold of which workflows there are. I would expect it to be possible in some way to run and debug the logic on my Ubuntu-laptop. Being able to run unit tests be nice...
I tried to compile for x86 as a target but the UGW-software suite from MaxLinear has no target for anything else than Mips and I could not find any emulator so that seems to be too much work.
I read through the but were not helped.
Then I found this
and started to think that maybe I can run OpenWRT in a Docker-container and in there install the few packages, currently netopeer2, libyang and sysrepo, I really need when testing the logic I'm working on. That would probably speed up things even though the best would be to be able to connect VSCode and gdb to the running process to be able to debug it.
I ran into trouble directly when starting the docker-image as seen below.

docker run --rm -it -v /tmp:/tmp openwrt/rootfs:mips_24kc
Unable to find image 'openwrt/rootfs:mips_24kc' locally
mips_24kc: Pulling from openwrt/rootfs
docker: no matching manifest for linux/amd64 in the manifest list entries.
See 'docker run --help'.

Looks fine until the "no matching manifest for linux/amd64 in the manifest list entries"

Any hints to increase efficiency and quality are very very welcome...