print-doc

name	print-doc
description	Print a document to the default printer. Accepts a markdown file (vault path or shorthand), a Google Docs URL, a Google Slides URL, a Google Drive file URL, or any local PDF. Markdown is rendered via pandoc + Chrome headless; Google Workspace files are exported as PDF via gws. Triggered by "print this", "print today's daily", "print [file]", "print this gdoc", "send to printer", or any request to put a doc on paper.
tags	["print","vault","obsidian","gdrive","utility"]
user_invocable	true

Print Document

Send a document to the system default printer. Supports markdown (vault notes), Google Docs, Google Slides, Google Drive files, and local PDFs — all collapsed to a single lp print call.

This skill assumes the printer is already set up in System Settings → Printers & Scanners. If your printer vendor's installer can't discover your Mac, the macOS AirPrint queue almost always works — add the printer there instead and skip the vendor app.

Prerequisites

• macOS (uses lp, sips, Chrome at the standard path)
• A default printer. Verify: lpstat -d. If no default, pass --printer <queue>.
• pandoc (brew install pandoc) — only required for markdown input
• Google Chrome at /Applications/Google Chrome.app/ — only required for markdown input. Chromium/Brave/Edge work too (swap the binary path in Step 2c).
• gws CLI configured — only required for Google Docs/Slides/Drive input
• jq — only required for the generic Drive branch (Step 4) — brew install jq
• For page-range printing: nothing extra; lp -P handles it

Input dispatch

The user supplies one of:

Input shape	Branch
today, yesterday, today's daily	Markdown → Step 2
<<VAULT_PATH>>/...md or any .md path	Markdown → Step 2
Bare filename or partial title	Search vault, then Markdown → Step 2
https://docs.google.com/document/d/<id>/...	Google Doc → Step 3
https://docs.google.com/presentation/d/<id>/...	Google Slides → Step 3
https://drive.google.com/file/d/<id>/...	Google Drive file → Step 4
Local .pdf path	Already-a-PDF → Step 5 directly

Optional flags (parse from the user's message):

Flag	Effect	lp option
-n N, N copies	print N copies	-n N
--duplex, double-sided, two-sided	double-sided (long-edge)	-o sides=two-sided-long-edge
--pages 1, --pages 1-3, --pages 2,4,7, page 1 only	restrict to a page range	-P <range>
--preview, preview first	open the PDF in Preview before printing; confirm with user	(skip lp)
--paper letter|a4	paper size (default: Letter)	-o media=Letter|A4
--printer <queue>	override default printer	-d <queue>

Step 1 — Setup temp workspace

Always work in a temp directory with named files. Do NOT use mktemp -t prefix.ext — on macOS that appends random chars after your template, so the file ends up without the expected extension, and Chrome will print HTML source as text instead of rendering it. Use mktemp -d and name the files yourself.

TMP=$(mktemp -d -t print-doc)
PDF="$TMP/out.pdf"

Step 2 — Markdown branch (pandoc + Chrome)

2a — Resolve file

case "$input" in
  today|"today's daily"*) file="<<VAULT_PATH>>/daily/$(date +%F).md" ;;
  yesterday|"yesterday's daily"*) file="<<VAULT_PATH>>/daily/$(date -v-1d +%F).md" ;;
  /*|~*) file="$input" ;;
  *) file=$(find <<VAULT_PATH>> -iname "*${input}*.md" | head -1) ;;
esac
[ -f "$file" ] || { echo "Not found: $file"; exit 1; }

If multiple matches, list and ask.

2b — Render to styled HTML

Pandoc consumes YAML frontmatter as metadata (won't print in body — good). Wikilinks [[Name]] and callout markers > [!note]- print as literal text — acceptable. To strip the callout markers: sed -E 's/> \[![a-z]+\]-? ?/> /' before pandoc.

TITLE=$(basename "$file" .md)
CSS="$TMP/style.css"
HTML="$TMP/doc.html"

cat > "$CSS" <<'EOF'
body { font-family: -apple-system, system-ui, sans-serif; max-width: 7in; margin: 0.5in auto; line-height: 1.5; color: #222; }
h1, h2, h3 { color: #111; }
h1 { font-size: 1.6em; border-bottom: 1px solid #ddd; padding-bottom: 0.2em; }
h2 { font-size: 1.3em; margin-top: 1.5em; }
code { background: #f4f4f4; padding: 2px 4px; border-radius: 3px; font-size: 0.9em; }
pre { background: #f4f4f4; padding: 0.8em; border-radius: 4px; overflow-x: auto; }
blockquote { border-left: 3px solid #ccc; padding-left: 1em; color: #555; margin-left: 0; }
table { border-collapse: collapse; margin: 0.5em 0; }
th, td { border: 1px solid #ddd; padding: 4px 8px; text-align: left; }
ul, ol { padding-left: 1.5em; }
hr { border: none; border-top: 1px solid #ddd; margin: 1.5em 0; }
EOF

pandoc "$file" -o "$HTML" --standalone --metadata title="$TITLE" -c "$CSS"

2c — HTML → PDF via Chrome headless

"/Applications/Google Chrome.app/Contents/MacOS/Google Chrome" \
  --headless --disable-gpu --no-pdf-header-footer \
  --print-to-pdf="$PDF" "file://$HTML" 2>/dev/null

Sanity-check: file "$PDF" should say PDF document. If it contains raw HTML text, the extension on $HTML was wrong — re-check 2b.

Browser alternates: swap the binary path for Chromium.app/Contents/MacOS/Chromium, Brave Browser.app/Contents/MacOS/Brave Browser, or Microsoft Edge.app/Contents/MacOS/Microsoft Edge.

Jump to Step 5.

Step 3 — Google Docs / Slides branch

Extract the file ID from the URL (docs.google.com/document/d/<id>/... or docs.google.com/presentation/d/<id>/...):

FILE_ID=$(echo "$url" | sed -E 's|.*/d/([^/]+).*|\1|')

Export directly as PDF — both Docs and Slides support application/pdf export:

gws drive files export \
  --params "{\"fileId\": \"$FILE_ID\", \"mimeType\": \"application/pdf\"}" \
  --output "$PDF"

Verify: file "$PDF" should say PDF document, version 1.4, N pages.

Jump to Step 5.

Step 4 — Google Drive file branch (generic)

For drive.google.com/file/d/<id>/..., first check the MIME type — could be a native PDF (use get), a Google Doc (use export), or something else.

FILE_ID=$(echo "$url" | sed -E 's|.*/file/d/([^/]+).*|\1|')
META=$(gws drive files get --params "{\"fileId\": \"$FILE_ID\", \"fields\": \"name,mimeType\"}")
MIME=$(echo "$META" | jq -r '.mimeType')

case "$MIME" in
  application/pdf)
    gws drive files get --params "{\"fileId\": \"$FILE_ID\", \"alt\": \"media\"}" --output "$PDF" ;;
  application/vnd.google-apps.document|application/vnd.google-apps.presentation)
    gws drive files export --params "{\"fileId\": \"$FILE_ID\", \"mimeType\": \"application/pdf\"}" --output "$PDF" ;;
  application/vnd.google-apps.spreadsheet)
    echo "Sheets-as-PDF rarely looks good; ask user to confirm before continuing"
    gws drive files export --params "{\"fileId\": \"$FILE_ID\", \"mimeType\": \"application/pdf\"}" --output "$PDF" ;;
  image/*)
    gws drive files get --params "{\"fileId\": \"$FILE_ID\", \"alt\": \"media\"}" --output "$TMP/img"
    sips -s format pdf "$TMP/img" --out "$PDF" ;;
  *) echo "Unhandled mime: $MIME"; exit 1 ;;
esac

Jump to Step 5.

Step 5 — Preview (if requested), then print

if [ "$preview" = true ]; then
  open -a Preview "$PDF"
  # ask user to confirm before continuing
fi

opts=()
[ -n "$copies" ] && opts+=(-n "$copies")
[ "$duplex" = true ] && opts+=(-o sides=two-sided-long-edge)
[ -n "$pages" ] && opts+=(-P "$pages")           # e.g. -P 1, -P 1-3, -P 2,4,7
[ "$paper" = "a4" ] && opts+=(-o media=A4) || opts+=(-o media=Letter)
[ -n "$printer" ] && opts+=(-d "$printer")

lp "${opts[@]}" "$PDF"

Report the job ID (e.g., request id is <queue>-NN) back to the user. Cancel with cancel <job-id>; check status with lpstat -W not-completed.

Notes

• Markdown frontmatter is metadata to pandoc — doesn't print in body.
• Wikilinks [[Name]] print verbatim.
• Embeds ![[file.png]] won't resolve.
• Daily-note dictation callouts print as quoted blocks — readable.
• Sheets as PDF: the export will fit each tab to a page; for long spreadsheets, ask the user if they want a specific tab or filtered range first.
• Slides: each slide becomes a page. --pages 1-5 for a deck excerpt.
• Long docs: default to --preview first if file "$PDF" reports >10 pages.
• Job tracking: lpstat -W not-completed. Clear stuck queue: cancel -a <queue>.