---
name: zlurp.ai
description: zlurp.ai provides a single web scraping skill that fetches a public URL and returns its content as clean markdown, along with metadata such as title, word count, and character count. It supports optional JavaScript rendering for dynamic pages. The service is designed for single-URL, on-demand content extraction.
host: zlurp.ai
---

# zlurp.ai

zlurp.ai is a focused web scraping host serving agents that need to convert public webpage content into structured markdown. It handles both static and JavaScript-rendered pages via a configurable mode and js flag. It does not offer bulk crawling, authenticated access, or multi-page workflows — it is a single-shot URL-to-markdown converter.

## When to use this host

Use zlurp.ai when an agent needs to extract readable text from a single public webpage and receive it as structured markdown. It is appropriate for both static HTML pages and JavaScript-rendered pages. Do not use it for pages behind login walls, paywalls, or any authentication barrier — those requests will fail. Do not use it for bulk or recursive site crawling; it processes one URL per call. If an agent needs to crawl multiple pages, index a site, or extract structured data at scale, a dedicated crawling or scraping platform with batch and session support would be more appropriate.

## Capabilities

### Web Content Extraction

Fetches a single public URL and returns its readable content as markdown, including page title and word and character counts, with optional JavaScript rendering support.

- **`scrape-url-to-markdown`** — Fetches any public URL and returns its content as clean markdown, with title, word count, char count, and scrape metadata.

## Skill reference

### `scrape-url-to-markdown`

**Zlurp Web Scraper** — Fetches any public URL and returns its content as clean markdown, with title, word count, char count, and scrape metadata.

*Use when:* Use when an agent needs to extract readable text content from a public webpage and receive it as structured markdown, including cases where the page may require JavaScript rendering.

*Not for:* Do not use for pages behind authentication or paywalls that require login credentials. Not suitable for bulk crawling or streaming content — this is a single-shot per-URL scrape.

**Inputs:**

- `url` (string, required) — The public URL to scrape. Must be a valid URI.
- `mode` (string) — Extraction mode: 'article' extracts main article content only; 'full' returns the entire page content. Defaults to 'article'.
- `js` (boolean) — Whether to render JavaScript before scraping. Set true for JS-heavy SPAs. Defaults to false.

**Returns:** Returns success=true with the page title, markdown body, word and char counts, and metadata flags for jsRendered, cachedResult, and scrapedAt timestamp.

**Example:** `{"url": "https://example.com", "mode": "article", "js": false}`

---