Skip to main content

Findkit Workers

Findkit Workers is a powerful way to modify how the crawler behaves. It allows arbitrary JavaScript code to be executed within the crawler. The API is inspired by Cloudflare Workers and uses Browser APIs such as the Fetch API and DOM.


A simple Findkit Worker looks like this:

// my-worker.js
// Workers are ESM modules with a default export
export default {
// The worker can handle one or more events
async fetch(request, context, next) {
// The fetch event allows you to hook into every http request the
// crawler sends out. The request and response objects implement
// the WHATWG Fetch Standard aka the Fetch API
console.log("Worker running in ", request.url);

// You can modify the request before it is sent
request.headers.set("Authorization", "***");

// The fetch event is handled as a middleware. Calling next() passes
// the request to the next worker and on the last worker the request is
// made and it returns the response.
const response = await next(request);

// Finally return the response. Possibly modifying it too.
// It is also possible to skip next() completely and to just return a
// custom response without making a network connection.
return response;

async html(page, context, next) {
// The html event is emitted for all 200 status text/html responses where
// the response html is parsed to DOM.

// The `page.window` property is like the `window` object in a browser
// but instead of being global it is just a local variable
const price = page.window.document.querySelector(".price")?.innerText;

const result = await next(page);

// Modify the result by adding a custom field
result.customFields.price = {
type: "number",
value: Number(price),

return result;

Testing and deployment

Add this to a my-worker.js file next to the findkit.toml file and add workers entry to the [[targets]] section:

id = "..."

host = ""
workers = ["my-worker.js"]

And now you can test the worker by running a test crawl

findkit crawl test ""

Once everything is ok, you can deploy it

findkit deploy

Add it will be executed within all crawl types: Full, partial and manual.

To remove the worker just remove the workers entry and run deploy again.


Most worker events are executed in a "middleware stack" meaning you can define multiple workers where the first worker passes the value to the next one using the next() function

workers = ["worker-a.js", "worker-b-js"]