Idea for optional reuse of custom chips across projects

[LeeLinkoff]

I’ve been building larger composite chips (adders, registers, small subsystems) and ran into a reuse limitation across projects that I wanted to share as an idea, mainly for advanced users.

Right now, custom chips are project scoped, which makes sense for determinism and learning. But once you start building higher level abstractions, it is common to want to reuse the same composite chips across multiple projects without manually copying them each time.

One idea I explored locally was an optional, explicit "global custom chip library", with these constraints:

• Completely opt in (default behavior unchanged)
• Projects remain fully self contained unless the user explicitly enables it
• Clear distinction between project chips and global chips
• When saving a custom chip, the user explicitly chooses:
  • save to project
  • save to global library
• Global chips are visually labeled so dependencies are never implicit

This would be aimed at advanced users only, without changing the NAND centric learning model or introducing new primitives. Beginners would not see or need this unless they opted in.

I am not assuming this belongs in the main build. I mainly wanted to share the idea in case others have run into the same reuse friction, or if it is something worth considering as an optional extension or documented pattern.

Totally fine if this is out of scope. Just wanted to put the idea out there.