Skip to content

Install

Get a running server and your first result in under a minute.

1. Install the binary

curl -fsSL https://olivierdevelops.github.io/webtasks/install.sh | sh

The installer detects your OS and architecture, downloads the matching webtasks binary, and drops it on your PATH. If no prebuilt binary is available for your platform it builds from source automatically (needs Go and git).

Other ways to install 
WEBTASKS_INSTALL_DIR=$HOME/bin \
  curl -fsSL https://olivierdevelops.github.io/webtasks/install.sh | sh

WEBTASKS_VERSION=v0.1.0 \
  curl -fsSL https://olivierdevelops.github.io/webtasks/install.sh | sh

git clone https://github.com/olivierdevelops/webtasks
cd webtasks
go build -o webtasks ./cmd/webtasks

One prerequisite: Chrome

webtasks drives an installed Chrome or Chromium — it doesn't bundle a browser. On a server or container, chromedp/headless-shell is the standard base image. ffmpeg is only needed for MP4 recording.

2. Get a bundle of tasks

A bundle is a folder of .webtask recipes the server exposes as endpoints. The repo ships a demo bundle with 38 example tasks — grab it to try things out:

git clone --depth 1 https://github.com/olivierdevelops/webtasks ~/webtasks

3. Start the server

WEBTASKS_BUNDLE=~/webtasks/demo webtasks

You should see:

[webtasks] starting on 127.0.0.1:8765
[webtasks] bundle: /home/you/webtasks/demo (dir)

Watch the browser

Add WEBTASKS_HEADLESS=false to open a visible Chrome window — handy while learning selectors.

WEBTASKS_HEADLESS=false WEBTASKS_BUNDLE=~/webtasks/demo webtasks

4. Check it's healthy

curl -s http://127.0.0.1:8765/health | python3 -m json.tool

{
  "ok": true,
  "pools": { "default": { "size": 3, "available": 3 } },
  "taskCount": 38
}

List every available task:

curl -s http://127.0.0.1:8765/tasks | python3 -m json.tool

5. Run your first task

curl -s -X POST http://127.0.0.1:8765/tasks/basics/title \
  -H 'Content-Type: application/json' -d '{}' | python3 -m json.tool

import requests
r = requests.post("http://127.0.0.1:8765/tasks/basics/title", json={})
print(r.json())

const r = await fetch("http://127.0.0.1:8765/tasks/basics/title", {
  method: "POST",
  headers: { "Content-Type": "application/json" },
  body: "{}",
});
console.log(await r.json());

{
  "ok": true,
  "data": {
    "page": {
      "title": "Example Domain",
      "heading": "Example Domain",
      "body": "This domain is for use in documentation examples …"
    }
  }
}

Pass inputs in the JSON body — for example a search query:

curl -s -X POST http://127.0.0.1:8765/tasks/search/duckduckgo \
  -H 'Content-Type: application/json' \
  -d '{"q":"golang"}' | python3 -m json.tool

6. Stream live progress (SSE)

Some tasks emit progress as they go. Ask for an event stream and watch it unfold:

curl -N -X POST http://127.0.0.1:8765/tasks/streaming/progress \
  -H 'Accept: text/event-stream' \
  -H 'Content-Type: application/json' \
  -d '{}'

SSE event stream

Next steps