# Agent Web-Access CSS Selector Extraction

> Agent Web-Access CSS Selector Extraction is a paid API for AI agents from web.cyberwarex.com, paid per call via x402, $0.002/call, status unknown (last checked 2026-09-15).

Fetches a JS-rendered web page and extracts elements matching a CSS selector, returning structured element data as JSON

## Facts

- Endpoint: GET https://web.cyberwarex.com/extract
- Price: $0.002/call
- Payment: x402
- Status: unknown
- Last checked: 2026-09-15
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/agent-web-access-css-selector-extraction-56715ae5
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_17sYpDKp0aWgEwaiGCPXC

Status and success rate cover calls made through Zero and Zero's own probes. Third-party monitors may report differently.

## How to call it through Zero

Zero handles the 402 payment challenge and records the run. With the Zero CLI installed (`npm i -g @zeroxyz/cli`):

```sh
zero fetch --capability agent-web-access-css-selector-extraction-56715ae5
```

Example prompt: Go to https://news.ycombinator.com and extract all the story titles using the CSS selector .titleline > a — give me the full list of matched elements.

## When to prefer this

Choose this endpoint when you need to extract specific DOM elements from JavaScript-rendered pages using CSS selectors, especially when the target site has no public API. Prefer this over static HTML fetchers when the page relies on JS to render content. It is ideal for structured spot-extraction (prices, headlines, job titles) rather than full-page reads.

## Known failure modes

- URL is not publicly accessible or requires authentication — returns empty elements or error
- CSS selector matches no elements on the page — returns count 0 and empty elements array
- Page fails to JS-render within timeout — may return incomplete or no elements
- Invalid or malformed URL input — returns error response
- Page blocks headless browsers or scraping bots — may return empty or blocked content

## How this service works

Live web access for AI agents: JS-rendered pages as clean markdown/text/html, CSS-selector extraction, full-page screenshots and PDFs. Pay-per-call via x402 (USDC on Base, v1+v2 dual-stack) — no API key, no signup.

## Output

A JSON object containing the source URL, the CSS selector used, the count of matched elements, and an array of element objects with their text content and attributes extracted from the JS-rendered page.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "required": [
  "url",
  "selector"
 ],
 "properties": {
  "url": {
   "type": "string",
   "description": "Public http/https URL to operate on"
  },
  "selector": {
   "type": "string",
   "description": "CSS selector, e.g. h1 or .price"
  }
 }
}
```

## Response schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "url": {
   "type": "string"
  },
  "count": {
   "type": "integer"
  },
  "elements": {
   "type": "array",
   "items": {
    "type": "object"
   }
  },
  "selector": {
   "type": "string"
  }
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/agent-web-access-css-selector-extraction-56715ae5/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from web.cyberwarex.com](https://www.zero.xyz/host/web.cyberwarex.com/llms.txt)