---

OpenClaw AI Agent Skill

---

name: web-reader description: Extract clean, readable markdown from any URL using Jina Reader (r.jina.ai). No API key needed for basic use. Use when you need to read a webpage as clean text, extract article content, scrape a page without JavaScript rendering, or get readable content from a URL that web_fetch struggles with.

Web Reader (Jina Reader)

Extract clean markdown from any URL via Jina Reader — no API key required.

Quick Start

scripts/read-url.sh "https://example.com/article"

How It Works

Jina Reader (r.jina.ai) renders any URL and returns clean markdown. Simply prepend https://r.jina.ai/ to any URL:

curl -s "https://r.jina.ai/https://example.com"

Advanced Options

With headers for better output

curl -s \
  -H "Accept: text/markdown" \
  -H "X-Return-Format: markdown" \
  "https://r.jina.ai/https://example.com"

Get just the main content (no nav/footer)

curl -s -H "X-With-Generated-Alt: true" \
  "https://r.jina.ai/https://example.com"

Search the web (Jina Search)

curl -s "https://s.jina.ai/your+search+query"

When to Use

• Use web-reader when web_fetch returns garbled/incomplete content
• Use web-reader for JavaScript-heavy pages (SPAs)
• Use web_fetch for simple pages (faster, no external dependency)
• Use browser tool for interactive pages requiring clicks/login

Rate Limits

• Free tier: ~20 requests/minute
• For heavy use, get an API key at jina.ai and pass via Authorization: Bearer <key>

Script

scripts/read-url.sh handles URL encoding and error checking.

Installation

cp -r web-reader/ ~/.openclaw/workspace/skills/web-reader/

License

MIT © Sentra Technology