codylrn804

Crawl4ai

資料分析v1.0.0
作者:codylrn804更新於 2026-05-17crawl4ai@clawhub_codylrn804/crawl4ai
網頁抓取#Web Scraping#Scrape

據 AlphaMoat 統計,截至 2026-10,Crawl4ai 累計下載 6,386 次、獲 2 個星標,按累計下載量在本次參與排名的 10,914 個 AI 技能中排名第 4,475 位,在「資料分析」分類(915 個)中排名第 419 位。

列表預設按累計下載量排序,也可按星標排序;詳情頁排名按累計下載量計算。

累計下載
6386
星標
2
全站下載量排名
#4,475
全站前 42% · 「資料分析」前 46%
平台累計下載星標評論
ClawHub ↗52552-
SkillHub ↗63862—

各平台的計數口徑不同,統計時間不一致。

這個技能是做什麼的

它結合傳統 HTML 解析與 AI,面向動態頁面、複雜 HTML、網頁內容清理和結構化等場景。

資料釋出於 2026-10-27

可以用來做什麼

  • 從網頁或 HTML 響應提取結構化資訊

什麼時候用

  • 頁面含動態內容或結構複雜,需要整理提取結果時

示例提示詞

以下是根據技能文件生成的示例,不代表實際執行結果。

從這個頁面提取產品名稱、文章內容和表格欄位

月度資料趨勢

累計下載6386
1579416167422026-03 · 累計下載 19352026-05 · 累計下載 30812026-07 · 累計下載 42402026-08 · 累計下載 47392026-09 · 累計下載 50202026-10 · 累計下載 638626-0326-0926-10
星標2
1232026-03 · 星標 22026-05 · 星標 22026-07 · 星標 22026-08 · 星標 22026-09 · 星標 22026-10 · 星標 226-0326-0926-10

2026-09 起統計方法調整:同名技能按作者分別統計,並覆蓋全部技能。圖中虛線為調整前的月份,與之後的資料不直接比較,也不計算跨越調整的增長。

月份累計下載環比星標
2026-106386—2
2026-095020—2
2026-084739+11.8%2
2026-074240—2
2026-053081—2
2026-031935—2

SKILL.md 原文

以下是作者釋出的原始檔案,按純文字顯示,本站未做修改或稽核;其中的連結不可點選,命令請核實後再執行。抓取於 2026-10-23。

7.6 KB · ClawHub v1.0.0
---
name: crawl4ai
description: AI-powered web scraping framework for extracting structured data from websites. Use when Codex needs to crawl, scrape, or extract data from web pages using AI-powered parsing, handle dynamic content, or work with complex HTML structures.
---

# Crawl4ai

## Overview

Crawl4ai is an AI-powered web scraping framework designed to extract structured data from websites efficiently. It combines traditional HTML parsing with AI to handle dynamic content, extract text intelligently, and clean and structure data from complex web pages.
展開全部(共 256 行)

## When to Use This Skill

Use when Codex needs to:
- Extract structured data from web pages (products, articles, forms, tables, etc.)
- Scrape websites with dynamic content or complex JavaScript
- Clean and normalize extracted data from various HTML structures
- Work with APIs or web services that return HTML
- Handle CORS limitations by scraping directly
- Process web content at scale with reliability

**Trigger phrases:**
- "Extract data from this website"
- "Scrape this page for [specific data]"
- "Parse this HTML"
- "Get data from [URL]"
- "Extract structured information from [website]"
- "Scrape [website] for [data type]"
- "Web scrape [URL]"

## Quick Start

### Basic Usage

```python
from crawl4ai import AsyncWebCrawler, BrowserMode

async def scrape_page(url):
    async with AsyncWebCrawler() as crawler:
        result = await crawler.arun(
            url=url,
            browser_mode=BrowserMode.LATEST,
            headless=True
        )
        return result.markdown, result.clean_html
```

### Extracting Structured Data

```python
from crawl4ai import AsyncWebCrawler, JsonModeScreener
import json

async def extract_products(url):
    async with AsyncWebCrawler() as crawler:
        result = await crawler.arun(
            url=url,
            screenshot=True,
            javascript=True,
            bypass_cache=True
        )
        # Extract product data
        products = []
        for item in result.extracted_content:
            if item['type'] == 'product':
                products.append({
                    'name': item['name'],
                    'price': item['price'],
                    'url': item['url']
                })
        return products
```

## Common Tasks

### Web Scraping Basics

**Scenario:** User wants to scrape a website for all article titles.

```python
from crawl4ai import AsyncWebCrawler

async def scrape_articles(url):
    async with AsyncWebCrawler() as crawler:
        result = await crawler.arun(
            url=url,
            javascript=True,
            verbose=True
        )
        # Extract article titles from HTML
        articles = result.extracted_content if result.extracted_content else []
        titles = [item.get('name', item.get('text', '')) for item in articles]
        return titles
```

**Trigger:** "Scrape this site for article titles" or "Get all titles from [URL]"

### Dynamic Content Handling

**Scenario:** Website loads data via JavaScript.

```python
from crawl4ai import AsyncWebCrawler

async def scrape_dynamic_site(url):
    async with AsyncWebCrawler() as crawler:
        result = await crawler.arun(
            url=url,
            javascript=True,  # Wait for JS execution
            wait_for="body",   # Wait for specific element
            delay=1.5,         # Wait time after load
            headless=True
        )
        return result.markdown
```

**Trigger:** "Scrape this dynamic website" or "This page needs JavaScript to load data"

### Structured Data Extraction

**Scenario:** Extract specific fields like prices, descriptions, etc.

```python
from crawl4ai import AsyncWebCrawler

async def extract_product_details(url):
    async with AsyncWebCrawler() as crawler:
        result = await crawler.arun(
            url=url,
            screenshot=True,
            js_code="""
                const products = document.querySelectorAll('.product');
                return Array.from(products).map(p => ({
                    name: p.querySelector('.name')?.textContent,
                    price: p.querySelector('.price')?.textContent,
                    url: p.querySelector('a')?.href
                }));
            """
        )
        return result.extracted_content
```

**Trigger:** "Extract product details from this page" or "Get price and name from [URL]"

### HTML Cleaning and Parsing

**Scenario:** Clean messy HTML and extract clean text.

```python
from crawl4ai import AsyncWebCrawler

async def clean_and_parse(url):
    async with AsyncWebCrawler() as crawler:
        result = await crawler.arun(
            url=url,
            remove_tags=['script', 'style', 'nav', 'footer', 'header'],
            only_main_content=True
        )
        # Clean and return markdown
        clean_text = result.clean_html
        return clean_text
```

**Trigger:** "Clean this HTML" or "Extract main content from this page"

## Advanced Features

### Custom JavaScript Injection

```python
async def custom_scrape(url, custom_js):
    async with AsyncWebCrawler() as crawler:
        result = await crawler.arun(
            url=url,
            js_code=custom_js,
            js_only=True  # Only execute JS, don't download resources
        )
        return result.extracted_content
```

### Session Management

```python
from crawl4ai import AsyncWebCrawler

async def multi_page_scrape(base_url, urls):
    async with AsyncWebCrawler() as crawler:
        results = []
        for url in urls:
            result = await crawler.arun(
                url=url,
                session_id=f"session_{url}",
                bypass_cache=True
            )
            results.append({
                'url': url,
                'content': result.markdown,
                'status': result.success
            })
        return results
```

## Best Practices

1. **Always check if the site allows scraping** - Respect robots.txt and terms of service
2. **Use appropriate delays** - Add delays between requests to avoid overwhelming servers
3. **Handle errors gracefully** - Implement retry logic and error handling
4. **Be selective with data** - Extract only what you need, don't dump entire pages
5. **Store data reliably** - Save extracted data in structured formats (JSON, CSV)
6. **Clean URLs** - Handle redirects and malformed URLs

## Error Handling

```python
async def robust_scrape(url):
    try:
        async with AsyncWebCrawler() as crawler:
            result = await crawler.arun(
                url=url,
                timeout=30000  # 30 seconds timeout
            )
            if result.success:
                return result.markdown, result.extracted_content
            else:
                print(f"Scraping failed: {result.error_message}")
                return None, None
    except Exception as e:
        print(f"Scraping error: {str(e)}")
        return None, None
```

## Output Formats

Crawl4ai supports multiple output formats:

- **Markdown**: Clean, readable text (`result.markdown`)
- **Clean HTML**: Structured, cleaned HTML (`result.clean_html`)
- **Extracted Content**: Structured JSON data (`result.extracted_content`)
- **Screenshot**: Visual representation (`result.screenshot`)
- **Links**: All links found on page (`result.links`)

## Resources

### scripts/
Python scripts for common crawling operations:
- `scrape_single_page.py` - Basic scraping utility
- `scrape_multiple_pages.py` - Batch scraping with pagination
- `extract_from_html.py` - HTML parsing helper
- `clean_html.py` - HTML cleaning utility

### references/
Documentation and examples:
- `api_reference.md` - Complete API documentation
- `examples.md` - Common use cases and patterns
- `error_handling.md` - Troubleshooting guide

「資料分析」分類下的熱門技能

全能金融爬蟲資料分析

中國金融機構資料爬取與分析綜合 Skill v12.0.0。2334家/43大類機構名單、60+財經媒體源輿情、 A股報告/公告/研報、60+種查詢多源金融資料(自動降級+快取)、爬取排程與網頁操作增強、 即時資料中樞、24 技術指標引擎、多源證據融合、110 MCP 工具 + 10 Resource + 14 Prompt。 🆕 v12.0.0 四條線(資訊獲取 / 打包下載 / 資料分析 / 登入): 編排:gather_info() 一次跑完"路由→抓取→去佔位→質量分→能力邊界→回執→打包", 8 類意圖各帶問法、輸入輸出、典型耗時與 not_for(說清哪件事做不了)。 誠實:反爬佔位條目一律剔除並記賬,空結果報 E_SOURCE_EMPTY,不把假資料判成 B 級成功。 交付:包內自帶 manifest.json(逐檔案 sha256),verify 查得出改動/缺失/多塞, 批次下載單條失敗不中斷整批,失敗清單可直接只重下那幾條。 登入:Cookie 由使用者匯出或掃碼接管,檔案化 + 健康度四檔,對外輸出全脫敏(值永不回顯)。 何時觸發:爬取機構、輿情爬取、東方財富、公告下載、即時行情、技術指標、統一查詢、 交付包、批次重下、我已經登入了、登入態還有效嗎(完整觸發詞見下方 auto_trigger)。

↓ 77萬★ 288user_9d5a2a39
鋼聯AI - 智慧問數資料分析

大宗商品價格資料查詢技能。用於查詢大宗商品期現貨價格、宏觀經濟指標、產業鏈供需資料、進出口庫存資料、金融市場資料等。當用戶詢問:(1) 期貨/現貨價格查詢,如"螺紋鋼價格"、"銅期貨收盤價"; (2) 歷史行情或趨勢分析;(3) 區域或企業價格對比;(4) 宏觀經濟指標,如GDP、CPI;(5) 產業鏈上下游資料;(6) 多市場金融資料(期權、股票、債券、外匯);(7) 結算確認單生成時觸發此技能。

↓ 70.5萬★ 6u_9ed3401d
smart-charts資料分析

32種圖表,3 個主題,2步生成: 1.上傳資料 — 將 CSV / Excel / JSON 等格式的資料檔案拖入對話方塊; 2.檢視結果 — 互動式圖表(HTML),可儲存圖表為圖片.

↓ 60.5萬★ 135user_5b28ea14
AdMapix資料分析

AdMapix 的取數層:把廣告與應用相關的查詢轉成介面呼叫,只回原始 JSON。

↓ 57.5萬★ 304fly0pants
論文配圖一鍵生成資料分析

論文配圖一鍵生成:一條命令從 JSON/CSV/Excel 生成頂刊級學術圖表,科研繪圖 與期刊配圖全場景覆蓋。40 種圖表(森林圖/KM 生存曲線(自動風險表)/競爭風險累計發生率/ROC/校準 曲線/決策曲線 DCA/列線圖/RCS 非線性劑量反應/亞組分析森林圖/RMST 限制平均生存時間/Galbraith 徑向圖/火山圖/雨雲圖/山脊圖/啞鈴圖/游泳圖(逐例時長+里程碑)/ 腫瘤緩解瀑布圖/UpSet 交集/PRISMA 流程圖/漏斗/BA 一致性/PCA/熱圖/箱線/小提琴/ 組合圖/Cox 多因素迴歸森林圖等),9 套配色(含色盲安全 Okabe-Ito 與期刊色板),9 種期刊投稿預設, YAML 流水線一條命令出全套論文圖,一次渲染多格式匯出(TIFF/PNG/PDF), PDF 重疊門禁,600dpi。純本地執行,資料不出機,中文零配置。適用:寫論文/ 投稿要配圖、把資料變成可投稿圖表、批次重出全套圖。不適用:文字潤色/翻譯/ 改寫(請用寫作類工具);商業報表/儀表盤/海報裝飾;無資料支撐的純示意圖。

↓ 42.1萬★ 83user_481397b5
smart-charts資料分析

Intelligent chart generation and data analysis skill. Reads user-supplied data files (CSV/Excel/JSON), analyzes data characteristics with LLM assistance, auto-recommends and generates interactive ECharts visualizations. Use when the user asks to analyze data, generate charts, create visualizations, or work with tabular data files.

↓ 25.7萬★ 1user_814dbe54
股票/基金/期貨量化分析工具資料分析

全球智慧投研工具(stock-researcher)v13.0:多源資料 + 機率預測 + 可匯出報告。 核心能力: - 多源資料路由:按市場/類別選源,逐級降級;每個結果標註來源與時效等級;源不可用時明確標 unavailable - 全資產量化:股票/ETF/場外基金/指數/期貨(期限結構、基差、持倉、折溢價)多因子分析 - 機率預測:1d/5d/1M/3M/6M 經驗分佈 + 分位數 + 方向機率 + 區間,kNN 鄰域收縮融合全域性分佈; 樣本不足自動棄權,可按歷史命中樣本(同桶)做機率校準 - 證據鏈簡報:每條結論繫結來源、時間與資料質量,衝突訊號保留而不抹平 - 報告匯出:Word/PPT/PDF/Markdown/JSON/純文字,純標準庫渲染 + 可選第三方引擎 - MCP Server 24 工具,Claude Code 可直接呼叫;知識庫 12 介面卡雙向同步 三條鐵律: - 資料不裝:拿不到就報不可用,不拿 0 價、不拿空表冒充 - 預測不裝:機率來自歷史相似條件的經驗分佈,樣本不足就說不足,不硬給方向 - 評分≠盈利保證:任何結論都帶免責宣告

↓ 19.2萬★ 147user_9d5a2a39
Tavily Search資料分析

透過 Tavily 搜尋公開網頁,並回傳含連結與摘要的結構化結果。

↓ 16.5萬★ 286Xiaoran Li

資料月份:2026-10 · 下載/星標數聚合自公開技能註冊庫(ClawHub、SkillHub),按月更新。口徑說明見資料方法。