Lets start simple. Paper works.
I dedicate one binder (rarely more) for each article-sized project. Each binder contains three main areas:
My first step in any project is to collect primary texts. These could be historical texts or contemporary text I am responding to. The idea is that they are the main objects of the project, the items I will engage with most directly. I try to be as spare as possible. Even if these "direct objects" come from larger works (say, Aristotle's Physics, when I care about Physics 2.2) I replicate them here. I will need to look back at these often, so simple is best.
The second step is to add relevant secondary literature. Whenever I read an article connected to the project, I add it here. If I read something once and didn't return to it for several months, I might remove it to save space, but only if I really need space.
The advantage of this system to me is that I have an easily browsable record of what the project involves. Yes, it involves paper. But the browsing ability is worth it. If I get stuck writing, it is sometimes helpful to leaf through the binder and remind myself where I've been. It can clarify where I'm going. This is especially useful if the project has been interrupted and I've stopped working on it for a while. Taking the binder out of the shelf is like a reintroduction, with out the need to remember what I read, where I left off, etc.
The long-term notes section is for written outlines, notes on the reading, etc. This section usually ends up being full of half-baked ideas I jotted down at some point, but I re-read them periodically to see what the project is about. Again, useful for being stuck. There is a danger here too: trying to fit too many half-baked ideas into one project will make it needlessly complex and even harder to write. Caveat emptor.
On the front, external page of the binder, I keep a table of contents. This makes browsing through the whole thing easier. Also, since I feel guilty about the paper output, it means that if I need to find a paper and I vaguely remember printing it out for some other project, I can just look at the front pages of binder and see if I have it. I'll have an easy template for this, which I'll provide in a future post.
Finally, student, I highly recommend that you keep track of seminars this way. I have binders from grad school that I can still trot out when I need a helping hand (e.g., How did Bernie Goldstein teach Newton? What did McDowell think is relevant background for Mind & World?). Also, putting something away on the shelf when you are really done with it is very satisfying.