I'd be wary of relying too much on this, although a mentor is obviously a good thing. Confidence is something you gain by experience, and experience is something you gain by just doing, so don't worry too much that you might not progress without someone to guide you - that kind of thinking really will hold you back.

It sounds to me (correct me if I'm wrong) that what you really need are non-trivial problems to solve, like you're (understandably) bored of doing things that don't really amount to anything (you can only write so many blogs and hello worlds). In which case I would start there, and find something to work on that gives you the stimulation you need. I know, I know, it's really tough to "come up with something" to work on, but one of the skills you'll develop is the ability to spot problems as well as solve them. Don't concern yourself with the size of the problem (but obviously don't bite off more than you can chew), the salient point is to start and finish things and gain your experience along the way. Each new problem yields entirely different lessons, so it's a worthwhile experience throughout your career - not just at the start :)

As far as finding projects is concerned, I'm just trying to recall some of the things I used to work on...

Here's one.. When I lived over in Kansas (I'm from the UK) I was really interested in the weather, and so I wrote a simple system that would pull data from the various weather stations around me and generated graphs as the weather fronts moved across the country. Not terribly useful for most people, but it solved a problem for me and taught me a lot about the weather (as well as how to deal with lots of datasets and generate useful reports and graphs from them).

And another.. A bunch of my friends & family and I play games together for which I host the servers on one of my dedicated servers. I wrote a utility that sits in the task bar and allows us to launch & connect directly to those servers rather than having to go into the server list etc., and also I wrote a bunch of things that hold the game servers together and allow me to create/configure/upgrade/etc (via Docker containers).

None of these things is particular complex, but each time I learn something new and the next time I come across a similar problem, I'm already partially equipped and my next solution will be better. Sometimes I even archive away my code and start afresh anyway, depending on whether I'm happy with what I've created (one of the stipulations I have is that it must be well written/designed enough that I can pick it up a year later and still work on it).

Edit: games... games are a good one. Are there games you play regularly for which you could create some useful web based tool or utility? Open sourced on Github would also give you some peer review.

Contributing to open source on GitHub is, probably, the best way of learning, learning from great codebases and learning from great engineers.

You also can use services like CodeMentor

Talking about the JavaScript, before doing any React stuff you should learn software engineering, architecture and JavaScript itself. Here I would be able to help and mentor you, and answer your questions since I would appreciate any help at BunnyJS, you may join Gitter and PM me your first direct questions, what would you like to know and what - to do.