TinyAgent Tutorial

Target

You’ll run one automation that visits a live store and returns the first two product names and prices as structured JSON — no selectors, no scraping rules, just a goal in plain English. Pick a tab below: through the Playground, or with the API.

Expected result

{
  "products": [
    { "name": "Bulbasaur", "price": "£63.00" },
    { "name": "Ivysaur", "price": "£87.00" }
  ]
}

Steps to do

Playground
API

Go to agent.tinyfish.ai and click Sign up — or Sign in if you already have an account. New accounts start with $8 of free usage, no credit card required.

Click Sign up in the top right of the homepage.

Create an account with GitHub, Google, Microsoft, X, or just an email address.

TinyFish create account screen with social sign-in options and an email field

Pick a sign-in method, or enter your email and click Continue.

Open the Agent Playground

After signing in you land on your console. Click Agent — either under Playground in the left sidebar, or the Agent card in the middle of the page.

TinyFish console home page with both Agent entry points highlighted: the sidebar link and the Agent card

Two ways in: the Agent link in the sidebar, or the Agent card below.

Describe your task

The Agent Playground asks what you want the agent to do.

Agent Playground with an empty goal field, ready for input

The goal box is where you describe the task, in plain English.

Type your goal, then Run

There’s one box, not separate URL and instruction fields — include the site right in the goal: Get to https://scrapeme.live/shop/ and extract the first 2 product names and prices. Return as JSON. Then click Run.

Agent Playground with the goal field filled in and the Run button highlighted

One box for the whole goal, URL included — then click Run.

Watch it run live

The Playground streams the browser live while the agent works, checking off each step of its plan as it completes it.

Agent run in progress, showing a live view of the browser navigating the target page

Watch the live browser view update in real time as the agent works.

Read the result

Once every step is checked off, the run is complete.

Completed agent run with every plan step checked off

Each plan step gets a checkmark as the agent finishes it.

The result appears below — as raw JSON, the same shape shown above:

Agent run result shown as raw JSON matching the expected output

Default view: the raw JSON result. Click the eye icon to switch views.

or toggle to a readable table for the same data:

Agent run result shown as a readable product table

Table view of the same result — handy for a quick human read.

Get an API key

Create one at API Keys, then store it in your environment:

export TINYFISH_API_KEY="your_api_key_here"

Call the TinyAgent API

from tinyfish import TinyFish

client = TinyFish()  # Reads TINYFISH_API_KEY from environment

with client.agent.stream(
    url="https://scrapeme.live/shop",
    goal="Extract the first 2 product names and prices. Return as JSON.",
) as stream:
    for event in stream:
        print(event)

import { TinyFish } from "@tiny-fish/sdk";

const client = new TinyFish();  // Reads TINYFISH_API_KEY from environment

const stream = await client.agent.stream({
  url: "https://scrapeme.live/shop",
  goal: "Extract the first 2 product names and prices. Return as JSON.",
});

for await (const event of stream) {
  console.log(event);
}

curl -N -X POST https://agent.tinyfish.ai/v1/automation/run-sse \
  -H "X-API-Key: $TINYFISH_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "url": "https://scrapeme.live/shop",
    "goal": "Extract the first 2 product names and prices. Return as JSON."
  }'

Read the result

The final COMPLETE event carries your result:

{"type":"COMPLETE","run_id":"abc123","status":"COMPLETED","result":{"products":[{"name":"Bulbasaur","price":"£63.00"},{"name":"Ivysaur","price":"£87.00"}]}}

Next Steps

How-to Guides

Structured output, multiple URLs, Vault logins, and Agent Mode

TinyAgent Reference

All endpoints across Agent, Runs, Profiles, and Vault

Runs

Understand status, polling, and completion

Goal prompting guide

Write more reliable automation instructions

Start Here

Search API

Monitor

Fetch API

Browser API

Research API

Datasets API

Capabilities

Tools & Integrations

Guides

Use Cases

Reference

TinyAgent Tutorial

Target

Steps to do

Next Steps

How-to Guides

TinyAgent Reference

Runs

Goal prompting guide

​Target

​Steps to do

​Next Steps

How-to Guides

TinyAgent Reference

Runs

Goal prompting guide

Target

Steps to do

Next Steps