A high-performance web content extractor that parses any URL into structured Markdown or plain text via the Zhipu AI Reader API.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install zhipu-web-fetch
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install zhipu-web-fetch using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Zhipu Web Page Reader?
The Zhipu Web Page Reader is a specialized utility designed for Openclaw Skills to bridge the gap between raw web URLs and structured, AI-ready data. By leveraging the Zhipu AI Reader API, this skill bypasses complex scraping logic and delivers clean Markdown or plain text versions of web pages, complete with metadata like titles and descriptions.
Whether you are building an AI agent that needs to read documentation or a researcher collecting data, this tool provides a lightweight, cURL-based solution. It ensures that the content fetched for your Openclaw Skills is properly formatted, stripping away unnecessary HTML noise while optionally retaining images and link summaries for contextual depth.
Zhipu Web Page Reader Use Cases
- Reading technical documentation or blog posts directly within an AI agent workflow.
- Extracting clean text for data analysis or LLM context windows.
- Converting complex web pages into GitHub Flavored Markdown for documentation syncing.
- Fetching web content without images to save bandwidth and token usage.
- Auditing page metadata, including SEO keywords and descriptions, across multiple URLs.
How Zhipu Web Page Reader Works
- The user provides a target URL and optional parameters such as return format or timeout.
- The skill initiates a POST request using cURL to the Zhipu AI Reader endpoint.
- The Zhipu API fetches the live web page and processes its DOM structure into a simplified format.
- The raw HTML is transformed into the requested format (Markdown or Text) using AI-driven parsing logic.
- A structured JSON object containing the parsed content and page metadata is returned to the Openclaw Skills environment.
Zhipu Web Page Reader Setup
To get started, ensure you have your API key and that curl is available on your system. Export your key to the environment:
export ZHIPU_API_KEY="your_api_key_here"
You can then use the provided wrapper script to fetch content for your Openclaw Skills projects:
# Basic fetch in Markdown format
bash scripts/zhipu_fetch.sh --url "https://example.com"
# Fetch plain text without images
bash scripts/zhipu_fetch.sh --url "https://example.com" --format text --no-images
Zhipu Web Page Reader Data Schema & Taxonomy
The skill returns a comprehensive response including the following data points:
| Field | Description |
|---|---|
reader_result.title |
The extracted title of the web page. |
reader_result.content |
The main body text formatted as Markdown or plain text. |
reader_result.description |
A concise meta description of the page content. |
reader_result.metadata |
An object containing keywords, viewport, and other HTML meta tags. |
reader_result.url |
The original URL that was processed. |
Zhipu Web Page Reader Advanced Features
- Cache Control: Use the no-cache flag to bypass Zhipu's internal caching for real-time data.
- Automated Summaries: Generate specific summaries for images and links to understand page structure without full rendering.
- Markdown Customization: Toggle GitHub Flavored Markdown (GFM) or choose to keep/discard image data URLs.
- Robust Timeouts: Configurable request timeouts (default 20s) to handle complex or slow-loading enterprise websites.
- Multi-format Support: Seamlessly switch between Markdown for documentation and plain text for simple data extraction.
SKILL.md
Loading
Related Openclaw Skills
Amap Location & Navigation Tool
A comprehensive CLI tool for Amap REST APIs providing weather, geocoding, and navigation services within the Openclaw Skills ecosystem.
aport-handoff
A skill designed to package completed work into verifiable handoff documents backed by signed APort decisions.
High School Geography Score Improvement
A specialized AI learning assistant designed to enhance geography exam scores through official textbook integration, error tracking, and expert-level problem analysis.
Architecture Designer
A high-level technical skill for designing scalable systems, making trade-off-informed decisions, and documenting them through ADRs.
Resume / CV Assistant
An AI-powered skill for polishing, tailoring, scoring, and exporting professional resumes and CVs.
System Prompts And Models Of Ai Tools
A comprehensive repository of internal system prompts, tool definitions, and model configurations for top-tier AI coding agents and development platforms.
Featured*
