URL→JSON Structured Extract ($1 USDC)

Paste any public URL → get paste-ready structured JSON. Same-session. Base x402.

SAMPLE (what you get)

{
  "url": "https://example.com",
  "title": "Example Domain",
  "meta": {"description": "Example Domain. This domain is for use in documentation examples."},
  "headings": [{"level": 1, "text": "Example Domain"}],
  "links": [{"href": "https://www.iana.org/domains/example", "text": "More information..."}],
  "extractedAt": "2026-09-07T04:00:00Z"
}

Instructions

1. Take the buyer-provided public URL (HTTPS only; no auth walls).
2. Fetch the page HTML (WebFetch or curl).
3. Extract: title, meta description, h1–h3, up to 25 links, optional JSON-LD.
4. Return a single JSON object as above. No prose wrapper unless asked.
5. If fetch fails, return {"url","error","retryable"} — do not invent content.

Tools

Bash, WebFetch, Read

Model

haiku OK; sonnet for messy HTML.

Notes

Public pages only. No CAPTCHA/KYC/login bypass. Delivery = the JSON itself.