Top 200 ReposEXPLORER
Browse agents, categories and pages
Back to results
Reported by sourceThe source says this happened. Top 200 Repos has not reproduced it.

Let an AI agent complete a multi-step task in a real browser

Describe a web task in plain language and let Browser Use drive a Chromium browser to finish it and return structured output.

Try the promptRead the original (opens in a new tab)

The problem

How it works

Browser Use is a Python library: you create an `Agent` with a task string, optionally pick an LLM such as OpenAI, Anthropic, Google or a local OpenAI-compatible server, and a local or cloud browser. The agent navigates, clicks and types to finish the task and returns a final result, with options for typed output, custom tools, file handling and domain restrictions.

Results and limitations

The README shows a demo of an agent booking a driving test slot and presents the library as a general way to automate browser work from a prompt.

Before you try it

The maintainers say domain limits are not a security sandbox, secrets passed in can still leak through screenshots or page scripts, CAPTCHAs are not guaranteed to work, and telemetry is on by default. Cloud browsers and models are billed separately. Results depend on your setup and were not independently reproduced by Top 200 Repos.

Tools and repositories

Tools
Python 3.11+ · uv or pip · Chromium
Prerequisites
An LLM API key or local model; BROWSER_USE_API_KEY only if you use the hosted browser and model.

Try it yourself: Run a supervised browser task

Fill the fields to adapt the prompt. Empty fields stay as visible [placeholders].

Preview
Write a Browser Use script that completes this task: [Task]. Restrict the agent to these domains: [Allowed domains], use [Model] as the model and a local Chromium browser. Return the result as JSON with these fields: [Output fields]. The agent must stop and ask me before submitting any form, logging in, accepting terms, purchasing, or sending anything, and must never handle passwords or payment data. Set ANONYMIZED_TELEMETRY=false. Show me the script and explain each setting before running it, then run it once and report what pages it visited.

Run browser agents in a container or throwaway profile and review every irreversible step yourself.

Related videos

About browser-use/browser-use

Independent creators. Listing a video implies no endorsement, sponsorship or partnership.

Sources

  1. browser-use/browser-use README (opens in a new tab)

    Project documentation · github.com · Undated · Source checked 9 Nov 2026, 22:22 UTC

    What this source supports
    • Agent takes a plain-language task and controls a browser
    • Supports OpenAI, Anthropic, Google and OpenAI-compatible local models
    • Python 3.11+; MIT-licensed library with a separate paid cloud
    • Domain restrictions are not a sandbox; run untrusted tasks in a container or VM
    • Telemetry enabled by default and can be disabled
Report a problem

Reported by source

Compare products across open tabs with Claude in Chrome, then buy yourself

Claude in Chrome can read and compare what is open in your tabs and pull details into a table; Anthropic’s own red-team numbers explain why it asks before purchases and why you should finish checkout.

claude.com · 25 Aug 2025
Reported by source

Fill out repetitive web forms with Skyvern's vision-based browser agent

Skyvern drives a browser with an LLM and computer vision to complete forms, such as quote requests or contact forms, on sites it has not seen before.

github.commicrosoft/playwright
Reported by source

Let an assistant test QA tickets in a browser via Playwright MCP

Peter Eysermans connected Claude Desktop to Playwright and GitHub MCP servers so it could walk through features from tickets, report results and draft reusable tests.

madewithlove.com · 22 Jul 2025microsoft/playwright
Reported by source

Prepare a tax return with an agent and compare it to an accountant

Kyle Corbitt worked through his 2025 US taxes with Codex, which researched rules, gathered documents and wrote a custom tax engine while a paid accountant did the same return.

corbt.com · 7 Mar 2026openai/codex, microsoft/playwright
Planned, not done

Try a desktop-operating agent safely in Anthropic's computer use demo container

Anthropic's computer-use-demo runs Claude inside a Docker desktop that you watch in the browser, with a chat panel, so you can test screen, mouse and keyboard automation away from your own machine.

github.com
Planned, not done

Delegate errands to Meta’s Muse, with approval before it sends or buys

Meta’s Muse runs in its own secure cloud VM, takes requests in WhatsApp or the Muse app, and asks before sending email or buying; a separate Sentinel agent must approve anything it sends to the internet.

about.fb.com · 8 Sept 2026