Remove API scripts, add pointer to SNflows/tendrils, continue re-organizing code.

[emirkmo]

https://github.com/SNflows/tendrils is the new API endpoint.

Now a discussion should be had about which API endpoints should actually be in there (e.g., should photometry be its own separate module?) Should query TNS/Alerce/etc. get some functionality there? (I think so).

After we decide, then the ones from here should be removed, and a pointer should be added to README and onto the wiki/website.

Finally, there are a bunch of scripts here that are not really part of the flows photometry pipeline, but are more helper scripts for e.g. populating the marshal. I propose that they should be moved to either flows-tools, or a separate module. The end goal is to isolate only the photometry and related pipeline scripts here, make it a version LOCKED module, and then release it as stable V1.0. Until then we should go to v0.10, etc.

• Move any remaining API-like scripts to tendrils (or decide where to put them, not in pipeline)
• Decide on modules for tendrils (should photometry API be its own?, TNS? Alerce?)
• Remove API scripts from flows pipeline, refactor code to reflect change.
• Move non-pipeline scripts to a separate repo or sub-module (recommend repo so we can easily make a pipeline package from the remaining and LOCK all dependencies). Maybe a marshal repo, or into flows-tools. Such as run_querytns, run_ztf, plot_lc, etc.
• Make V1.0 of photometry pipeline with as few dependencies as possible and with locked versions. Release as package.

[emirkmo]

The linked PR moves the API wrapper to Tendrils.

To make even less dependencies, some of the Runners should really be moved to a separate repo, probably flows-tools for now. (Such as run-querytns, run-ztf, etc.)

Eventually, they would make sense in a flows Webserver/Database repo. This repo can also have the python based API (ReST) that interacts with the database.