web-browsing-manual

Name: web-browsing-manual
Availability: InStock
Author: huangzesen

Retrieve web content across PDFs, APIs, static HTML, and JS-rendered pages.

621|58|Updated Mar 15, 2026

Search Tags:#data-extraction #arxiv #crossref #openalex #web-browsing #tiered-strategy

Ready to run

npx skills add https://github.com/huangzesen/lingtai --skill web-browsing-manual

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: web-browsing-manual
Source: https://github.com/huangzesen/lingtai/tree/main/tui/internal/preset/skills/web-browsing-manual
Command: npx skills add https://github.com/huangzesen/lingtai --skill web-browsing-manual

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires requests, beautifulsoup4, and includes scripts (resource) and assets (resource) components.

What problem does it solve?

This Skill provides a structured, four-tier approach to retrieving and extracting information from the web, reducing manual browsing time and ensuring consistent results.

Core Features & Use Cases

Four-tier strategy: Tier 0 Direct PDF access; Tier 1 metadata APIs (arXiv, CrossRef/OpenAlex); Tier 2 structured extraction from static HTML; Tier 3 JS-rendered or login-gated pages via Playwright stealth.
Site playbooks and templates: Google Scholar, Nature, Springer, arXiv, PubMed, and NASA ADS, with bundled site templates and CSS-selector guidance.
Bundled assets and scripts: assets/api-endpoints.json, assets/site-templates.json, assets/css-selectors.json, assets/regex-patterns.json, and a runnable extract_page.py to drive auto-tiering and fallbacks.

Quick Start

Provide a URL and the tool will auto-select a tier and fetch data accordingly.

Frequently Asked Questions about web-browsing-manual

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I extract academic metadata from arXiv, CrossRef, and OpenAlex automatically?▼

You can extract academic metadata from arXiv, CrossRef, and OpenAlex by applying the tiered strategy that queries metadata APIs directly, using bundled endpoint configurations and regex patterns to fetch structured results without manual browsing.

What is the best way to scrape static HTML and JS-rendered pages for academic articles?▼

The best way to scrape static HTML and JS-rendered pages is using a four-tier approach that attempts static HTML extraction first, then falls back to Playwright stealth for JavaScript-rendered or login-gated content using predefined CSS selectors.

Does this web browsing tool support direct PDF retrieval from Nature, Springer, and PubMed?▼

Yes, this web browsing tool supports direct PDF retrieval from Nature, Springer, and PubMed through its Tier 0 direct access strategy, utilizing bundled site templates and CSS-selector guidance to locate and download full-text documents.

Can I use BeautifulSoup and requests to scrape Google Scholar and NASA ADS results?▼

Yes, you can use BeautifulSoup and requests to scrape Google Scholar and NASA ADS results because the tool includes predefined site playbooks and bundled scripts that drive automated tiering and fallbacks for these specific academic platforms.

How do I automate web content extraction with fallback logic for different site structures?▼

You can automate web content extraction with fallback logic by providing a URL to the bundled extract_page.py script, which auto-selects the appropriate tier and applies site-specific templates and regex patterns to ensure consistent retrieval.

Why does structured data extraction fail on some academic publisher websites?▼

Structured data extraction fails on some academic publisher websites when pages are JavaScript-rendered or login-gated, requiring a fallback to Tier 3 Playwright stealth browsing instead of standard static HTML parsing with BeautifulSoup.