FAQ
Common questions about runtime, indexing, and current limitations.
FAQ
Does eye index every repository automatically?
No. Indexing is lazy and query-triggered. There is no watch mode or background daemon.
What runtime does index-backed navigation require?
Universal Ctags on PATH, plus ripgrep for discovery and fallback search.
What happens when the cache schema changes?
Old .eye/cache.db and .eye/blobs/ state is invalidated instead of migrated in place. The next index-backed operation rebuilds it.
Can name-based lookups still be ambiguous?
Yes. Name-only lookups can return multiple candidates, and some anchor or reference lookups can fall back to lower-confidence text search when ctags output is not enough.
Is the server optimized for dumping entire files?
No. context is bounded and designed for navigation, not whole-definition dumping.
Where do heavy validation fixtures live?
The default local validation flow stays lightweight. Large OSS coverage lives in tests/fixtures/real/, and the heavy suite runs separately with pnpm run test:fixtures:real.