Import

You can bring existing documents into the Knowledge Base instead of retyping them — one file at a time, or a whole tree of Markdown at once. Use Import from the Knowledge Base.

Supported files

File	What you get
.md / .txt	One article from the file’s text
.docx (Word)	One article — the text is extracted to Markdown
.zip	Bulk import — many articles, see below

The text is imported, and images embedded in the document are carried over as article attachments (see Embedded images below). The original file itself is not stored, and non-image binaries are not kept.

When you import, you choose:

• a Category (the home folder the imported article(s) land in), and
• a Status — import as Draft (private to you) or Published.

Import is an article-write action and runs as the importing person, never as a service account.

Single file

Pick a .md, .txt or .docx file, choose a category and status, and Import. The file is processed and you are taken straight to the new article. Larger files take a moment.

Bulk import from a .zip

A .zip lets you migrate an existing wiki — for example a folder of Markdown notes, or an exported Obsidian or Notion vault — in a single upload.

• Only .md and .txt entries are imported, along with the archive’s folder structure: nested folders in the zip are recreated as folders in the Knowledge Base, so the hierarchy comes across. Anything else (images, .docx inside the zip, hidden files, binaries) is skipped, not treated as an error.
• [[slug]] wiki-links inside the imported notes are reconnected to the freshly created articles where possible, so a cross-linked vault arrives already wired together rather than as a wall of dead links. A link that still has no target degrades to the usual “not created yet” mention.
• Name clashes are resolved automatically. If an imported article’s slug is already taken, it is given a numbered suffix (-2, -3, …) rather than failing. Every rename is reported.

Reading the results

A bulk import runs in the background and, when it finishes, shows a per-item summary so nothing is silent:

• Created — articles imported cleanly.
• Renamed — imported, but the slug was auto-suffixed to avoid a clash (the requested name is shown).
• Skipped — entries that were not imported, with the reason (a folder entry, a hidden file, an unsupported type, or an empty file).

The summary also reports how many folders were created and how many wiki-links were reconnected. Review it, then close the dialog — a bulk import does not navigate you away to a single article.

Limits and safety

• There is a maximum file size for imports (a server setting; the default is small). An oversized file is rejected up front with a clear message.
• Archives are unpacked in a sandboxed, memory-capped process, so a malicious or accidentally enormous .zip (a “decompression bomb”) cannot exhaust the server — it fails that one import safely and leaves everything else running. A failure of this kind is permanent for that file: fix or shrink the archive and import again.

Embedded images

Images embedded inside an imported file are now brought across automatically, so a migrated Word or Markdown runbook keeps its screenshots:

• A picture pasted into a .docx, or a base64 image embedded directly in Markdown, is extracted, saved as an article attachment, and shown inline in the imported article — the same way an image you paste into the editor is (see Articles and authoring). This also applies to .md entries inside a .zip.
• What is not brought across: images linked from the web (https://…) are left as links and not shown (lazyit never fetches remote images); separate image files sitting alongside notes in a .zip are skipped; and drawings that aren’t real raster images — SVG or HTML — are not imported. Export those to PNG first.
• Each image passes the same checks as an editor upload: its true type is verified, it is re-encoded (stripping camera/location metadata), and it counts against your instance’s attachment storage limit. An image that can’t be read is dropped from the article; if attachment storage is full, the whole import fails with a clear message — free space (or ask your administrator to raise the limit) and import again. Documents with a very large number of embedded images keep only the first several dozen.

References you write yourself to images already uploaded in lazyit (attachment: links) are kept as-is.