Archiving Clipping Media

Skip: wikilinks (![[...]]), text hyperlinks, already-local paths (../assets/...), data: URLs.

Download Commands

Images/direct files:

curl -L --fail -o "content/private/assets/{filename}" "{url}"

YouTube/Vimeo:

# Get the actual video title first for the filename
yt-dlp --get-title "{url}"
yt-dlp --merge-output-format mp4 -o "content/private/assets/{filename}" "{url}"

CDN URLs — try stripped URL first, fall back to original if it 404s:

• Economist: remove cdn-cgi/image/width=...,quality=...,format=.../ prefix
• General: strip query params like ?w=600&quality=80
• If the stripped URL fails (404/403), retry with the original CDN URL

Filename Convention

Pattern: {article-slug}-{descriptive-name}.{ext}

• article-slug: kebab-case, first few meaningful words of the title
• descriptive-name: from the URL's last path segment or alt text
• ext: preserve original extension; use .mp4 for yt-dlp downloads
• Check content/private/assets/ for collisions before saving

Examples from existing assets: ane-m4-hero.jpeg, ane-software-stack.png, rodney-terminal-demo.jpg

Replacement Format

CRITICAL: Use standard markdown with relative paths. Do NOT use Obsidian wikilink embeds.

Images — preserve any alt text:

![alt text](../assets/article-slug-name.png)

HTML video tags — replace entire tag with markdown:

![](../assets/article-slug-video.mp4)

HTML audio tags — replace entire tag with markdown:

![](../assets/article-slug-audio.mp3)

YouTube embeds — same pattern after yt-dlp download:

![](../assets/article-slug-video-title.mp4)

Error Handling

• 404/failed download: report and leave original link intact
• Large files (>50MB): report size, ask user before keeping
• Duplicate URLs: download once, replace all occurrences

Post-Download

1. ls -lh each downloaded file to verify non-empty
2. Read modified clipping to confirm replacements
3. Report summary: downloaded count, failed count, total size

Common Mistakes