Skill

web-clipper

Clip a web page into a clean Markdown file under raw/, with images downloaded locally to raw/attachments/. Self-contained — depends only on the defuddle npm package installed inside this skill folder. Use when given a URL and asked to capture it for the vault without firecrawl or API keys.

npx claudepluginhub jasonsie/zkfy --plugin zkfy

Tool Access

This skill uses the workspace's default tool permissions.

Preview

Fetch a URL → `raw/<slug>.md` with all images materialized to `raw/attachments/<slug>/`.

Supporting Assets

package-lock.jsonpackage.jsonscripts/clip.sh

SKILL.md

Similar Skills

using-superpowers

180.9k

Mandates invoking relevant skills via tools before any response in coding sessions. Covers access, priorities, and adaptations for Claude Code, Copilot CLI, Gemini CLI.

3 files

superpowers

Stats

Stars0

Forks1

Last CommitApr 28, 2026

Actions

View Source View Plugin View on GitHub View README

Help us improve

Share bugs, ideas, or general feedback.

Web Clipper

Fetch a URL → raw/<slug>.md with all images materialized to raw/attachments/<slug>/.

Input

$ARGUMENTS — <url> [--no-attachments] [--out <path>]

<url> (required) — the page to clip
--no-attachments — skip image download; keep remote URLs as-is
--out <path> — override default output path (raw/<slug>.md)

Setup (one-time)

cd skills/web-clipper && npm install

This installs defuddle (+ its optional deps linkedom and turndown) locally into skills/web-clipper/node_modules/. No API key or external service required.

Runtime requirements: node, curl, jq.

Execution

Run the wrapper script from the vault root:

bash skills/web-clipper/scripts/clip.sh <url>
bash skills/web-clipper/scripts/clip.sh <url> --no-attachments
bash skills/web-clipper/scripts/clip.sh <url> --out raw/my-note.md

Pipeline

URL
 │
 ▼
[1] defuddle CLI  (skills/web-clipper/node_modules/.bin/defuddle)
    defuddle <url> --markdown --json
    → JSON: title, author, published, domain, contentMarkdown
 │
 ▼
[2] Compose Markdown file
    • YAML frontmatter  (Date, Type, Link, Categories, tags, Before, Next, …)
    • # <title> heading
    • contentMarkdown body
    → raw/<slug>.md
 │
 ▼
[3] Attachment materialization  (skipped with --no-attachments)
    • grep remote image URLs from output
    • curl → raw/attachments/<slug>/<md5>.<ext>
    • rewrite Markdown links to local relative paths

Frontmatter produced

---
Date: <today>
Type: literature
Link: '<url>'
Categories: []
Sub-Categories: []
Aliases: ['<page title>']
tags: []
Before: ''
Next: ''
Author: '<author>'       # omitted if empty
Published: '<date>'      # omitted if empty
Source: '<domain>'       # omitted if empty
---

Compatible with /zk-note and /source-to-zk downstream.

Output layout

raw/
  <slug>.md
  attachments/
    <slug>/
      <md5>.png
      <md5>.jpg
      …

Continue the pipeline

/zk-note raw/<slug>.md
# or
/source-to-zk raw/<slug>.md

Verification

# Sanity: defuddle installed
skills/web-clipper/node_modules/.bin/defuddle --help

# Clip with attachments
bash skills/web-clipper/scripts/clip.sh https://react.dev/learn

# Clip without attachments
bash skills/web-clipper/scripts/clip.sh https://react.dev/learn --no-attachments

# Pipe into the ZK pipeline
/source-to-zk raw/react-dev-learn.md