u2j ▸ tool opened 0

Turn any URL into structured JSON

Paste a page you are authorized to access, name the fields, and get clean JSON back - with copy and curl. No login.

Writ runs on your own accounts, with your own credentials and data, on sites you are authorized to use.
run ▸ extract to json 3

This is a preview that shapes example output from your inputs. A live extraction runs inside the product on workflows. Use it only on pages you are authorized to access.

dif ▸ preview to api 7
01
the differentiators

The same extraction, callable on demand

This sandbox shapes example output. In Writ, the extraction becomes a real endpoint your software runs - even behind logins and TOTP on your own authorized accounts.

A REST endpoint and an MCP tool

Publish the extraction once and it becomes a JSON API your software calls, plus an MCP tool your AI agents can invoke directly.

Runs on a schedule or on change

Trigger it on a schedule, or the instant a monitored page changes - monitors check as fast as every 10s on higher plans and fire in seconds.

Self-healing selectors

When a page shifts, Writ heals the target so runs keep working. Cloud AI can rewrite the recipe when a run fails; local, BYO-key runs are unbilled.

seq ▸ how it works 11
02
how it works

From a page to structured data

One-off here. Repeatable in the product.

01

Paste a URL

Enter the address of a page you have the right to access.

02

Pick what to extract

A whole-page summary, or a few fields you name.

03

Get clean JSON

Structured output you can copy, or grab as a curl command.

04

Make it repeatable

Publish the same extraction as a Writ endpoint your software calls on demand.

rep ▸ do it on demand 15
03
on demand in writ

Do it once here. Do it on demand in Writ.

Define the extraction once and publish it as a workflow. It becomes a REST endpoint and an MCP tool your software and AI agents call - and it can run when a monitored page changes.

q ▸ common questions 19
04
faq

Questions about URL → JSON

What does the tool return?
A clean JSON object representing the page - its title, key fields, and the structured data you ask for. It is a preview of the kind of typed output Writ produces, designed to be copied or turned into a curl command.
Can I extract specific fields?
Yes. Name the fields you want - like price, title, and availability - and the output is shaped to match. In the full product you define an extraction schema once and reuse it on every run.
Is this allowed?
Use it only for pages you have the right to access. Writ frames web data extraction as something you do on sites you are authorized to use. Writ runs on your own accounts, with your own credentials and data, on sites you are authorized to use.
How do I make this run on a schedule or on demand?
Publish the extraction as a Writ workflow. It becomes a REST endpoint and an MCP tool you can call from your software or an AI agent, run on a schedule, or trigger when a monitored page changes.
go ▸ turn it into an api 23
05
get started

Make extraction repeatable

Publish a page as a callable JSON API your software can hit on demand. Start free.

Get started See workflows