defuddle

Defuddle

Defuddle extracts clean article content from a web URL — stripping navigation, ads, sidebars, and other clutter. It is the right tool for turning a single web page into a clean Markdown note.

---

Installation

npm install -g defuddle-cli

Check if already installed:

defuddle --version

---

Basic Usage

Extract as Markdown (recommended)

defuddle parse <url> --md

Save to file

defuddle parse <url> --md -o content.md

Get a specific metadata property

defuddle parse <url> -p title
defuddle parse <url> -p domain
defuddle parse <url> -p author
defuddle parse <url> -p published
defuddle parse <url> -p description

Output as JSON (includes all metadata + content)

defuddle parse <url> --json

Output as HTML

defuddle parse <url> --html

---

Integration Pattern: Web Article to Vault Note

The fastest pattern for archiving a single web article into Matt's Vault:

1. Extract content with defuddle
2. Wrap in frontmatter using the obsidian-markdown skill's standard template
3. Write to vault using obsidian-cli

# Step 1: get title and content
TITLE=$(defuddle parse "https://example.com/article" -p title)
defuddle parse "https://example.com/article" --md -o /tmp/article.md

# Step 2: prepend frontmatter (use Write tool, not echo)
# Step 3: create in vault
obsidian create name="$TITLE" vault="Matt's Vault" content="$(cat /tmp/article.md)"

---

When to Use Defuddle vs Other Tools

Use case	Tool
Single web article → vault note	defuddle
Entire documentation site (Apple HIG, library docs)	doc-harvest plugin
Web page that requires JavaScript rendering	WebFetch tool (browser-based)
Already have clean text, just need to write to vault	obsidian-cli skill directly

---

Notes

• Defuddle works best on article-style pages (blog posts, documentation pages, news articles).
• It does not execute JavaScript — purely HTML parsing. SPA sites that require JS rendering will yield empty or minimal content. Use WebFetch for those.
• For SPA-rendered documentation sites (e.g., Apple Developer docs), the doc-harvest plugin has specialized handling via JSON API discovery.