Skip to content

Importing your world

Waymark can pull existing notes and documents into memory cards without losing its memory-first identity. Every import is explicit and preview-first:

  • Folders are never scanned automatically — you always name the file or folder.
  • Each imported file becomes a sources row and a sourced memory, so retrieval can cite it (for example pdf: brief.pdf).
  • The same resolved file path is skipped on re-import unless you pass --force.

Supported formats

Format Suffixes How text is extracted Dependency
Markdown .md, .markdown Heading + first paragraph none
Plain text .txt, .text Raw text none
PDF .pdf Existing text layer, no OCR optional pypdf
Word .docx Paragraph text (no styles/images) none (stdlib)

No OCR

PDF import reads the existing text layer only. Scanned or image-only PDFs report "no extractable text" rather than running OCR.

Import a single file

waymark import markdown .\notes\idea.md
waymark import text .\notes\standup.txt
waymark import pdf .\docs\brief.pdf
waymark import docx .\docs\weekly.docx

For the binary formats you can inspect the extracted card before saving:

waymark import pdf .\docs\brief.pdf --preview
waymark import docx .\docs\weekly.docx --preview

If pypdf is not installed, install the PDF extra:

pipx inject waymark-memory "waymark-memory[pdf]"

Use py -m pipx inject ... on Windows or python3 -m pipx inject ... on Linux if pipx is not on PATH yet.

Import a whole folder

waymark import folder previews every supported file in a folder, then imports them only when you add --apply:

waymark import folder .\dropbox                 # preview only
waymark import folder .\dropbox --apply         # import the previewed files
waymark import folder .\dropbox --apply --recursive --limit 100

Legacy Markdown folder command

waymark import markdown-folder still works for older scripts, but new workflows should use waymark import folder. The unified command covers Markdown plus text, PDF, and DOCX files with the same preview-first safety model.

A batch never aborts on a bad file:

  • Files whose path was already imported are reported as duplicates.
  • Unreadable files (or PDFs with no extractable text) are reported as skipped.
  • Everything else still imports.

Because imported paths are skipped, re-running the same command resumes naturally — already-imported files are left alone.

Guided import

In the guided interface (waymark), the Import screen offers the same single-file buttons and a Preview Folder → Apply Preview flow for the multi-type folder import.

After importing

waymark sources list          # what came from where
waymark ask "the topic"       # answers cite imported sources