Very good evening, and thank you Laurent et al for all the work on the excellent Joplin and Joplin Cloud.
I'm tinkering with a system to move data in and out of Joplin from scripts and at the terminal, and have a couple of questions.
A. State of the CLI npm package
I started installing the Joplin CLI, almost as described at Joplin Terminal Application | Joplin , but was somewhat alarmed to see 25+ npm warnings about deprecated packages when I left out --loglevel=error.
npm audit also reports a decent number of issues.
What would be the advisable option for a developer wanting to use the CLI?
- Install just as suggested. (It will be fine, there are no actual issues, say.)
- Install as suggest, but run
npm update,npm audit fix, and friends. - Pull the Git repo and run the cli app from
main. Pull from the remote regularly. - Something else?
B. Is joplin-cli the best option for my use cases?
I want to set up a system where I can automatically pull and push Markdown text from arbitrary locations in the file system, to/from Joplin. Ideally in a fully headless, containerized environment, with no display server.
It seems the CLI app in batch mode would be a good choice to move data in and out of Joplin's db. But, should I look at other options? Can I use the celebrated Data API without actually running the desktop app?
All the very best, and thanks for any help, suggestions or corrections 
Further general context: I use the Android app, and sync to Joplin Cloud. I use E2EE. For the use cases discussed here, I would use a container running some Debian or Fedora image.
The purported use cases: I have a somewhat strange bespoke Markdown-to-tex-transpiler that I want to run on Joplin notes, with the aim of rendering print-ready pdfs. I also have a Markdown-to-epub script. All these are self-standing Python utilities. These tools work much like running latexmk in watch mode, say.