🕸️ Zion Dependency Graph
See every reference. Trust every delete.
Performance & Large Projects
Initial scan
The first scan iterates project assets and builds both dependency and reverse-reference indexes. On typical projects this takes seconds; large projects can take longer.
Cache is saved to Library/ZionDependencyGraph/cache.json and reused between sessions.
Code reference scan
The string-reference scan only reads .cs files plus three lookup tables, so it is quick even on large projects. It runs cooperatively in the background (a few milliseconds per editor frame) and updates incrementally when scripts are saved.
Duplicate detection is opt-in
Hashing size-colliding files can take tens of seconds on texture-heavy projects, so it runs only when you click Find Duplicates - never automatically on editor start, script compiles, entering Play mode, or after a rescan. Deleting duplicates updates the list in place without re-hashing.
Graph rendering cap
Graph view caps displayed nodes (200) to preserve editor responsiveness. If exceeded, output is truncated and status indicates it.
Cached analyses
Orphan/heavy computations are memoized by cache version and recomputed after rescan.