Skip to main content
Pick the task you’re trying to accomplish. Each guide assumes you already know the basics from the Tutorial, goes straight to the code, and ends with a Common Mistakes section for the failure modes people actually hit.

Run TinyAgent From ChatGPT or Claude

Connect the plugin once, then trigger automations from a normal chat

Extract Pricing as Structured JSON

Pin down an exact, predictable result shape — e.g. product name and price

Run Parallel Agents Across Multiple Sites

Fire the same goal at many URLs at once, e.g. comparing deals across sites

Build a Google Form With a TinyProfile

Reuse a saved, signed-in browser session so the agent skips the login step

Scrape Behind a Login With Vault Credentials

Let the agent log in with credentials from your connected password manager

Fill Out and Submit a Form, Including a Job Application

Complete a multi-step form behind a login wall, credentials in Vault not the goal

Test Workflows With Agent Configurations

Use Strict mode and step limits so a test run fails fast instead of wandering

Troubleshoot Agent Setup and Failed Runs

Find out why a request was rejected or a run came back without your data

Recover a Stuck or Long-Running Run

Tell a slow run from a stuck one, stop it, and split the work so the next run finishes

Compare Prices Across Websites

Check one product or quote on several sites at once, every price back as JSON

Test Your App on localhost or a Preview URL

Describe a flow in plain English and get back pass or fail as JSON