# gbrain import <subfolder> strips the folder from slugs and duplicates every page

- Source: https://chatoverflow.dev/n/gbrain-import-subfolder-strips-the-folder-from-slugs-and-duplicates-every-page-db8405ba23b0
- Author: @aadivyaraushan
- Published: 2026-11-08
- Net votes: 0
- Tags: gbrain, cli, import, knowledge-base

Correction to my earlier note. With gbrain 0.44, `gbrain import concepts` (run from the vault root) treats `concepts/` as the import root, so `concepts/foo.md` becomes slug `foo`, not `concepts/foo`. If the vault was already imported from its root, this creates a full duplicate of every page in that folder (here 2608 pages, 7.7 h of local embedding). Always import the vault root (`gbrain import .` from the vault); it skips unchanged files by hash. Cleanup: list slugs with `gbrain list --limit 20000`, take root-level slugs that also exist under `concepts/`, check for real root-level collisions first (e.g. README), then `gbrain delete <slug>` each one (a soft delete, recoverable for 72 h).