Ready to supercharge your AI agents?

Install any skill in seconds. Free to start, no credit card required.

Cheerio · terminalskills | DecimalAI

terminalskills/cheerio

SkillScore96Maintained

Parse and extract data from HTML with Cheerio. Use when a user asks to scrape static web pages, parse HTML files, extract data from HTML, build a web scraper for server-rendered pages, extract text or links from HTML documents, parse RSS/XML feeds, transform HTML content, or process HTML emails. Covers jQuery-style selectors, DOM traversal, text extraction, attribute parsing, and integration with HTTP clients for web scraping pipelines.

codingby terminalskills

Links this skill to your agents and keeps it current with the author’s updates.

Automatic

Coding pack

1 · One-time — install the CLI (docs ↗)

pip install decimalai

2 · Get the skill — writes .claude/skills/terminalskills-cheerio/SKILL.md

decimalai skills pull terminalskills-cheerio --out .claude/skills/

No CLI? Get the same file — curl · raw file

curl -fsSL --create-dirs -o .claude/skills/terminalskills-cheerio/SKILL.md https://app.decimal.ai/s/terminalskills-cheerio/SKILL.md

Raw file: /s/terminalskills-cheerio/SKILL.md ↗ · machine-readable summary at /s/terminalskills-cheerio.json ↗

In productionGlobal

how this skill behaves on real traffic · last 30 days

No production data for this skill yet — usage numbers appear here once agents start running it.

Anonymized aggregate across every org using this skill — proof from real production use, not a demo.

Does this skill help? · Benchmark: with vs. withoutsign in to run this on your agents

25 test cases from the publisher’s eval.yaml · each run twice on gemini-3.6-flash — with the skill injected, then without · last run 55d ago

Yes — +4% lift · at higher cost+143% tokens0% turns96% of whole cases pass with skill (24/25)

Test case	Without → With	Effect	Δ tokens	Δ turns
case-03	✗→✓	▲ Improved	72%	0%
case-01	✓→✓	＝ Same ✓	42%	0%
case-02	✓→✓	＝ Same ✓	53%	0%
case-05	✓→✓	＝ Same ✓	157%	0%
case-06	✓→✓	＝ Same ✓	386%	0%

Showing 5 of 25 cases

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v4 — 8/1/2026

live_secretremote_code_executionreverse_shelldata_exfiltrationssrf_metadataagent_snoopinganti_refusaldestructive_commandsobfuscated_unicodeunicode_homoglyphopaque_blobsuspicious_urltool_overreachtrigger_abuseprompt_injection_phrasing

IntentSafe

LLM judge · does the skill do what it claims

ContentClean

LLM judge · policy & harmful-content review

Source

Imported

terminalskills / skills

skills/cheerio/SKILL.md

View on GitHub ↗

License Apache-2.0Imported 8/1/2026Version v1

Skill contents

SKILL.md

Cheerio

Overview

Cheerio is a fast, lightweight HTML/XML parser for Node.js that implements a jQuery-like API. Unlike Puppeteer, it does not run a browser — it parses raw HTML strings, making it 100x faster and ideal for scraping server-rendered pages, parsing HTML files, and transforming HTML content. Pair it with fetch or axios for web scraping, or use it standalone for HTML processing.

Instructions

Step 1: Installation

bash
npm install cheerio

Step 2: Parse HTML and Extract Data

javascript
// parse_html.js — Load HTML and extract structured data with CSS selectors
import * as cheerio from 'cheerio'

const html = `
<html>
  <body>
    <h1>Products</h1>
    <div class="product" data-id="1">
      <h2>Widget Pro</h2>
      <span class="price">$29.99</span>
      <a href="/products/widget-pro">Details</a>
    </div>
    <div class="product" data-id="2">
      <h2>Gadget Max</h2>
      <span class="price">$49.99</span>
      <a href="/products/gadget-max">Details</a>
    </div>
  </body>
</html>`

const $ = cheerio.load(html)

// Extract all products
const products = []
$('.product').each((i, el) => {
  products.push({
    id: $(el).attr('data-id'),
    title: $(el).find('h2').text().trim(),
    price: $(el).find('.price').text().trim(),
    link: $(el).find('a').attr('href'),
  })
})

console.log(products)
// [{ id: '1', title: 'Widget Pro', price: '$29.99', link: '/products/widget-pro' }, ...]

Step 3: Web Scraping with Fetch

javascript
// scrape_site.js — Fetch a page and extract data
import * as cheerio from 'cheerio'

async function scrape(url) {
  const response = await fetch(url)
  const html = await response.text()
  const $ = cheerio.load(html)

  // Extract all links
  const links = []
  $('a[href]').each((i, el) => {
    links.push({
      text: $(el).text().trim(),
      href: $(el).attr('href'),
    })
  })

  // Extract meta tags
  const meta = {
    title: $('title').text(),
    description: $('meta[name="description"]').attr('content'),
    ogImage: $('meta[property="og:image"]').attr('content'),
  }

  return { links, meta }
}

Step 4: Advanced Selectors and Traversal

javascript
// selectors.js — Complex CSS selectors and DOM traversal
const $ = cheerio.load(html)

// Attribute selectors
$('a[href^="https"]')           // links starting with https
$('img[src$=".png"]')           // PNG images
$('div[class*="product"]')      // divs with "product" in class

// Traversal
$('.product').first()            // first product
$('.product').last()             // last product
$('.product').eq(2)              // third product (0-indexed)
$('.price').parent()             // parent of each .price element
$('.product').children('h2')     // direct h2 children
$('.product').find('.price')     // descendants matching .price
$('.product').next()             // next sibling
$('.product').prev()             // previous sibling

// Filtering
$('.product').filter((i, el) => {
  const price = parseFloat($(el).find('.price').text().replace('$', ''))
  return price < 50
})

// Text and HTML
$('.product').first().text()     // all text content, flattened
$('.product').first().html()     // inner HTML

Step 5: Table Extraction

javascript
// extract_table.js — Parse HTML tables into structured data
function extractTable($, tableSelector) {
  /**
   * Convert an HTML table to an array of objects using headers as keys.
   * Args:
   *   $: Cheerio instance
   *   tableSelector: CSS selector for the table element
   */
  const headers = []
  $(`${tableSelector} thead th`).each((i, el) => {
    headers.push($(el).text().trim())
  })

  const rows = []
  $(`${tableSelector} tbody tr`).each((i, tr) => {
    const row = {}
    $(tr).find('td').each((j, td) => {
      row[headers[j]] = $(td).text().trim()
    })
    rows.push(row)
  })
  return rows
}

// Usage
const tableData = extractTable($, '#pricing-table')
// [{ Plan: 'Free', Price: '$0', Users: '1' }, { Plan: 'Pro', Price: '$29', Users: '10' }]

Step 6: HTML Transformation

javascript
// transform.js — Modify HTML content
const $ = cheerio.load(html)

// Add class
$('.product').addClass('featured')

// Remove elements
$('.ad-banner').remove()

// Replace content
$('h1').text('Updated Title')

// Wrap elements
$('.product').wrap('<section class="product-section"></section>')

// Add attributes
$('a').attr('target', '_blank')
$('img').attr('loading', 'lazy')

// Get modified HTML
const modifiedHtml = $.html()

Examples

Example 1: Build a price monitoring scraper

User prompt: "Scrape product prices from 5 competitor websites daily and save to a CSV. The sites are server-rendered (no JavaScript needed)."

The agent will:

Use fetch + cheerio for each site (no browser overhead).
Write site-specific selectors for product name, price, and availability.
Parse prices into numbers, normalize currency.
Append results to a CSV with timestamps.
Set up as a cron job for daily execution.

Example 2: Extract and clean article content from HTML

User prompt: "I have 1,000 saved HTML pages from a blog. Extract just the article title, author, date, and body text from each, ignoring navigation, ads, and footers."

The agent will:

Read each HTML file, load with cheerio.
Extract content using article-specific selectors (article, .post-content, etc.).
Strip HTML tags from body, normalize whitespace.
Output structured JSON with title, author, date, and clean text.

Guidelines

Use cheerio for server-rendered pages (where the HTML contains the data you need). For SPAs or JavaScript-rendered content, use Puppeteer instead.
Cheerio does not execute JavaScript, fetch external resources, or render CSS — it only parses the HTML string you give it.
Always call .trim() on extracted text — HTML often contains whitespace, newlines, and indentation that clutters results.
Use .attr('href') and .attr('src') to get link/image URLs. Remember these may be relative — resolve them against the base URL.
For large-scale scraping, cheerio is 100x faster than Puppeteer and uses negligible memory. It can process thousands of pages per second.
Combine cheerio with fetch or axios for scraping, and add delays between requests to avoid overwhelming target servers.

Illustrative Want to see what terminalskills/cheerio would put in front of your model? Preview the rendered body and check its trigger phrases — no LLM call, no signup.

Latest benchmark · publisher runVerified

8/4/2026, 04:40 UTC · 25 cases

Lift

+4%

96% vs 92%

Agent turns

1.0 vs 1.0

Tokens

+143%

2,388.08 vs 982.84

fail → pass · 1pass → pass · 23fail → fail · 1

Per-case results (25)

Case	Status	Duration (ms)			Turns			Tokens			Tool calls
Case	Status	Without	With	Δ	Without	With	Δ	Without	With	Δ	Without	With	Δ
case-01 I need to extract product prices from a web application built as a single-page React client where the initial HTTP fetch response only returns `<div id="root"></div>` and all data is rendered dynamically via client-side JavaScript. What Node.js approach should I use?	pass→pass	13,737	8,550	-38%	1	1	0%	2,396	3,400	+42%	0	0	—
case-02 I am writing a Node.js script that needs to log into a vendor portal by filling input forms, clicking submit buttons, and waiting for AJAX responses in a real browser environment. Which tool category or library fits this workflow best?	pass→pass	9,869	5,750	-42%	1	1	0%	1,805	2,758	+53%	0	0	—
case-03 I am writing frontend JavaScript code running inside user web browsers to handle button clicks and modify client DOM elements on page load. Should I use a server-side HTML string parser library in my browser application bundle for this?	fail→pass	9,151	6,960	-24%	1	1	0%	1,722	2,956	+72%	0	0	—
case-04 I need to parse 50,000 static server-rendered HTML pages per minute in Node.js without launching headless browsers or executing embedded JavaScript. What lightweight library and approach offers the highest speed and lowest memory overhead for string parsing?	fail→fail	17,093	9,898	-42%	1	1	0%	3,194	3,716	+16%	0	0	—
case-05 When parsing header titles and price tags from HTML strings in Node.js using Cheerio, extracted text strings often include unwanted leading whitespace, tabs, and line breaks. What specific JavaScript string method should be chained onto extracted text values to clean them up?	pass→pass	4,634	2,487	-46%	1	1	0%	874	2,245	+157%	0	0	—
case-06 When looping over hyperlinked elements in a loaded Cheerio object `$`, what Cheerio method call extracts the target link URL from an `<a>` tag?	pass→pass	2,354	2,174	-8%	1	1	0%	442	2,150	+386%	0	0	—
case-07 When extracting link attributes from HTML with Cheerio using `.attr('href')`, many values return relative paths like `/catalog/item-42`. What step must be performed on these relative paths before making subsequent network requests?	pass→pass	4,752	3,798	-20%	1	1	0%	845	2,455	+191%	0	0	—
case-08 Write a Cheerio CSS attribute selector string to match all anchor tags whose `href` attribute starts with `https`.	pass→pass	2,724	2,373	-13%	1	1	0%	508	2,092	+312%	0	0	—
case-09 Write a Cheerio CSS attribute selector string to match all image elements whose `src` attribute ends with `.png`.	pass→pass	2,218	1,808	-18%	1	1	0%	351	2,037	+480%	0	0	—
case-10 Write a Cheerio CSS attribute selector string to match all `div` elements whose `class` attribute contains the substring `product`.	pass→pass	2,700	2,603	-4%	1	1	0%	516	2,190	+324%	0	0	—
case-11 In a Cheerio document `$`, there are multiple matching `.product` elements. What zero-indexed method call selects specifically the third matching `.product` element?	pass→pass	3,335	2,131	-36%	1	1	0%	600	2,034	+239%	0	0	—
case-12 Given a Cheerio selection of `.price` elements, what method retrieves the direct parent element of each `.price` element?	pass→pass	2,642	2,377	-10%	1	1	0%	448	2,158	+382%	0	0	—
case-13 Given a selection of `.product` elements in Cheerio, what method finds all nested descendant elements with the class `.price`?	pass→pass	3,169	2,592	-18%	1	1	0%	522	2,153	+312%	0	0	—
case-14 In Cheerio, how do you navigate from a `.product` element to its immediate previous sibling element and immediate next sibling element?	pass→pass	5,566	3,343	-40%	1	1	0%	1,122	2,391	+113%	0	0	—
case-15 In Cheerio, you want to filter a set of `.product` elements to retain only those elements meeting a custom numerical price threshold logic. What Cheerio traversal method accepts a predicate callback function to filter elements?	pass→pass	4,339	3,153	-27%	1	1	0%	687	2,269	+230%	0	0	—
case-16 In Cheerio, what is the key functional difference between `.text()` and `.html()` when called on an element?	pass→pass	7,028	5,840	-17%	1	1	0%	1,283	2,779	+117%	0	0	—
case-17 When extracting tabular data from an HTML table `#pricing-table` using Cheerio, what specific CSS selector string targets the header cell elements inside the table head section?	pass→pass	2,832	3,043	+7%	1	1	0%	499	2,144	+330%	0	0	—
case-18 When transforming tabular HTML rows (`tbody tr`) into an array of JavaScript objects using Cheerio, how are individual data cell values (`td`) matched to their corresponding table column header names?	pass→pass	9,845	6,550	-33%	1	1	0%	2,017	2,989	+48%	0	0	—
case-19 In a loaded Cheerio document `$`, what method adds the CSS class `featured` to all elements matching `.product`?	pass→pass	1,834	1,611	-12%	1	1	0%	276	1,946	+605%	0	0	—
case-20 In Cheerio, what method completely removes all DOM elements matching `.ad-banner` from the document?	pass→pass	2,948	2,681	-9%	1	1	0%	526	2,144	+308%	0	0	—
case-21 In Cheerio, what method call replaces the text contents of all `h1` elements with 'Updated Title'?	pass→pass	2,208	1,804	-18%	1	1	0%	365	2,020	+453%	0	0	—
case-22 In Cheerio, what method call wraps each matching `.product` element inside `<section class="product-section"></section>`?	pass→pass	2,649	1,250	-53%	1	1	0%	502	1,891	+277%	0	0	—
case-23 In Cheerio, how do you add `target="_blank"` to all link tags and `loading="lazy"` to all image tags?	pass→pass	5,448	3,396	-38%	1	1	0%	1,208	2,383	+97%	0	0	—
case-24 In Cheerio, after performing various DOM manipulations on `$`, what method exports the full updated HTML string?	pass→pass	4,810	2,769	-42%	1	1	0%	782	2,154	+175%	0	0	—
case-25 In Cheerio, how do you extract the `content` attribute from the Open Graph meta tag `<meta property="og:image" content="...">`?	pass→pass	5,233	2,813	-46%	1	1	0%	1,081	2,248	+108%	0	0	—

These cases were also screened by the skill’s intent / safety pass.See the Trust review ↗Reproduce this benchmark in Colab ↗

How this skill was measured

DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 25 cases were attempted. The headline lift of +4 percentage points is the difference between those two pass rates over the 25 comparable cases.

Model measured: gemini-3.6-flash
Graded by: gemini-3.6-flash
Grading method: judged
Verification: verified
Pass rate without the skill: 92%
Pass rate with the skill: 96%
Cases attempted: 25
Cases counted toward lift: 25
Cases that regressed: 0
Skill version measured: v1
Run completed: 8/4/2026, 04:40 UTC

Measured example — same prompt, with vs. without the skill · from the benchmark run · case case-03

Prompt

I am writing frontend JavaScript code running inside user web browsers to handle button clicks and modify client DOM elements on page load. Should I use a server-side HTML string parser library in my browser application bundle for this?

Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.

Show both answers

SkillScore blends 4 signals1 of 4 complete

BenchmarkPass rate WITH skill on a verified run96

Live evalNeeds ≥10 evaluated traces in 30 days

AI ratingNeeds ≥10 rated traces

AdoptionNeeds ≥10 serves across ≥2 orgs in 30 days

Efficiency · separate axis (at equal correctness)+0% turns+170% tokens

Shown for transparency; not credited unless a verified run shows no correctness regression.

Backed by 1 of 4 signals — each added signal makes the score more robust.

Verified on gemini-3.6-flash · re-tested 55 days ago

Other measured skills in the registry, with their headline benchmark lift.