My team at work works in agile as well. One of the things I recently realized I was doing was keeping “monolithic” projects in OF and trying to organize tasks around them, like projects that didn’t change very often, because they would take multiple sprints to complete. So my project list used to look like:
* Website X
* Website Y
* Website Z
While that kept my project list short, it didn’t really serve me well as a bucket of tasks, because the list of tasks that looking at a project would generate would be all over the place (many contexts, many states), and relating to a myriad of different deliverables.
Also, in GTD speak, this meant that I was using OF projects more as Areas of Focus (20,000 foot) instead of as Projects (10,000 foot).
Now, I have significantly changed my approach to have many more projects that open/close much more frequently, but that relate to much more specific commitments/outcomes. My project list now looks like:
* Build the template for Page A of Website X
* Classify/plan outstanding requests from the business
* Refactor CSS for Website Z
In this way, the projects are not just buckets of tasks that kind of relate to each other, but are actual completable things, and I give myself permission to create and close projects daily (though I make most of them as part of sprint planning at the top of a sprint).
I have an on hold context called Someday/Maybe that I can use as a backlog, though I do keep a few other lists (single action projects) as support for my workflow:
* Project Concepts (held as tasks, but represent potential future projects)
* Properties (again, held as tasks, but actually a reviewable list of what was my project list)
So, while I haven’t really answered your question, @lgrassini, I hope my experience with the problem might help think of the problem in other ways?
Cheers,
ScottyJ