### Local Development Setup

Source: https://github.com/browser-use/browser-use/blob/main/skills/open-source/references/quickstart.md

Local development setup: clone the repository, sync dependencies with uv, and use helper scripts for setup, linting, testing, and running examples.

```bash
git clone https://github.com/browser-use/browser-use
cd browser-use
uv sync --all-extras --dev

# Helper scripts
./bin/setup.sh   # Complete setup
./bin/lint.sh    # Formatting, linting, type checking
./bin/test.sh    # CI test suite

# Run examples
uv run examples/simple.py
```

--------------------------------

### Install browser-harness with uv

Source: https://github.com/browser-use/browser-use/blob/main/skills/qa/SKILL.md

Installs browser-harness using uv, a one-time setup that takes about 30 seconds. If uv is missing, the command installs it first via the official installer. After installation, verifies that browser-harness is on PATH.

```bash
command -v uv || curl -LsSf https://astral.sh/uv/install.sh | sh   # the uv installer, only if missing
uv tool install "git+https://github.com/browser-use/browser-harness"
command -v browser-harness                                         # verify it's on PATH now
```

--------------------------------

### Setup and run basic task

Source: https://github.com/browser-use/browser-use/blob/main/examples/cloud/README.md

Sets up the environment and runs the basic Cloud API V4 example. Requires an API key added to the .env file.

```bash
uv sync
cp examples/cloud/env.example .env
# Add your API key to .env
uv run python examples/cloud/01_basic_task.py
```

--------------------------------

### Run helper scripts for setup, linting, and testing

Source: https://github.com/browser-use/browser-use/blob/main/AGENTS.md

Run the helper scripts for common development tasks: setup.sh installs uv, creates a venv, and installs dependencies; lint.sh runs all pre-commit hooks; test.sh runs the core test suite used in CI.

```bash
# Complete setup script - installs uv, creates a venv, and installs dependencies
./bin/setup.sh

# Run all pre-commit hooks (formatting, linting, type checking)
./bin/lint.sh

# Run the core test suite that's executed in CI
./bin/test.sh
```

--------------------------------

### Run the quickstart script

Source: https://github.com/browser-use/browser-use/blob/main/examples/integrations/anthropic/README.md

Executes the quickstart script with uv.

```bash
uv run run_browser.py
```

--------------------------------

### Install required packages

Source: https://github.com/browser-use/browser-use/blob/main/examples/integrations/slack/README.md

Installs the required packages for the Slack integration example.

```sh
pip install fastapi uvicorn slack_sdk
```

--------------------------------

### CLI prompt for installing and connecting browser-use

Source: https://github.com/browser-use/browser-use/blob/main/README.md

Paste this prompt into Claude Code, Codex, Hermes, OpenClaw, or your favorite agent to install or upgrade browser-use, register the skill, and connect it to your browser. If setup or connection fails, the prompt directs you to the browser-harness install guide.

```text
Install or upgrade browser-use to the latest stable version with uv using Python 3.12, run `browser-use skill install` to register the skill, and connect it to my browser. If setup or connection fails, follow https://github.com/browser-use/browser-harness/blob/main/install.md.
```

--------------------------------

### Run the simple example

Source: https://github.com/browser-use/browser-use/blob/main/AGENTS.md

Run the simple example script using uv.

```bash
uv run examples/simple.py
```

--------------------------------

### Install browser-use with uv

Source: https://github.com/browser-use/browser-use/blob/main/skills/open-source/SKILL.md

Install the browser-use library using uv, the recommended package manager. Run 'uv pip install browser-use' followed by 'uvx browser-use install' to set up the library.

```bash
uv pip install browser-use
```

```bash
uvx browser-use install
```

--------------------------------

### Set up project and install dependencies

Source: https://github.com/browser-use/browser-use/blob/main/examples/integrations/anthropic/README.md

Initializes a Python 3.12 project, adds the browser-use and anthropic packages, and installs the browser-use CLI.

```bash
uv init --python 3.12
uv add browser-use anthropic
uvx browser-use install
```

--------------------------------

### Install browser-use with uv

Source: https://github.com/browser-use/browser-use/blob/main/README.md

Install Browser Use with uv. Requires Python >= 3.11; run `uv init --python 3.12` first if starting a new project.

```bash
uv add browser-use
```

--------------------------------

### Clone and install Browser Use

Source: https://github.com/browser-use/browser-use/blob/main/AGENTS.md

Clone the repository and install dependencies using uv or pip. The uv command syncs all extras and dev dependencies; the pip alternative installs directly from the main branch.

```bash
git clone https://github.com/browser-use/browser-use
cd browser-use
uv sync --all-extras --dev
# or pip install -U git+https://github.com/browser-use/browser-use.git@main
```

--------------------------------

### Manual session creation and usage

Source: https://github.com/browser-use/browser-use/blob/main/skills/cloud/references/sessions.md

Create a session with profile, proxy, and start URL. Run multiple tasks in the same session, get the live URL for monitoring, and stop the session to save state.

```python
session = await client.sessions.create(
    profile_id="uuid",           # Persistent profile
    proxy_country_code="us",     # Residential proxy
    start_url="https://example.com",
)

# Run multiple tasks in same session
await client.run("First task", session_id=session.id)
await client.run("Follow-up task", session_id=session.id)

# Get live URL for monitoring
session_info = await client.sessions.get(session.id)
print(session_info.live_url)  # Watch agent in real-time

await client.sessions.stop(session.id)
```

--------------------------------

### Install eth-account (Python)

Source: https://github.com/browser-use/browser-use/blob/main/skills/x402/SKILL.md

Install the eth-account library to generate a new Ethereum wallet in Python.

```bash
pip install eth-account
```

--------------------------------

### Name Actions Directly: Task Prompt Example

Source: https://github.com/browser-use/browser-use/blob/main/skills/open-source/references/agent.md

Demonstrates naming actions directly in the task prompt to guide the agent's behavior.

```python
task = """
1. Use search action to find "Python tutorials"
2. Use click to open first result in a new tab
3. Use scroll action to scroll down 2 pages
4. Use extract to extract the names of the first 5 items
"""
```

--------------------------------

### Check Browser Use CLI with --doctor and skill show

Source: https://github.com/browser-use/browser-use/blob/main/skills/remote-browser/SKILL.md

Run these commands to verify the Browser Use CLI is installed and working. If setup fails, follow the linked Browser Use skill.

```bash
browser-use --doctor
browser-use skill show
```

--------------------------------

### Install viem (TypeScript)

Source: https://github.com/browser-use/browser-use/blob/main/skills/x402/SKILL.md

Install the viem library to generate a new Ethereum wallet in TypeScript.

```bash
npm install viem
```

--------------------------------

### Install browser-use CLI

Source: https://github.com/browser-use/browser-use/blob/main/skills/cloud/references/guides/tools-integration.md

Install the browser-use CLI package using uv pip. This is the first step before using any CLI commands.

```bash
uv pip install 'browser-use[cli]'
```

--------------------------------

### Install Python SDK

Source: https://github.com/browser-use/browser-use/blob/main/skills/cloud/references/quickstart.md

Install the browser-use-sdk package for Python.

```bash
pip install browser-use-sdk
```

--------------------------------

### Quick Example: Custom Tool with BrowserSession

Source: https://github.com/browser-use/browser-use/blob/main/skills/open-source/references/tools.md

Shows the minimal setup for a custom tool: create a Tools instance, decorate an async function with @tools.action, and pass the tools object to the Agent. The parameter must be named browser_session: BrowserSession (not browser: Browser) because the agent injects by name matching; a wrong name fails silently.

```python
from browser_use import Tools, ActionResult, BrowserSession

tools = Tools()

@tools.action('Ask human for help with a question')
async def ask_human(question: str, browser_session: BrowserSession) -> ActionResult:
    answer = input(f'{question} > ')
    return ActionResult(extracted_content=f'The human responded with: {answer}')

agent = Agent(task='Ask human for help', llm=llm, tools=tools)
```

--------------------------------

### Install browser-use

Source: https://github.com/browser-use/browser-use/blob/main/examples/apps/ad-use/README.md

Install the latest version of browser-use, which includes screenshot functionality required by this demo. Requires browser-use v0.7.7 or higher.

```bash
pip install -U browser-use
```

--------------------------------

### Start ngrok to expose local server

Source: https://github.com/browser-use/browser-use/blob/main/examples/integrations/slack/README.md

Starts ngrok to expose your local server. Replace 3000 with the port your server runs on.

```sh
ngrok http 3000
```

--------------------------------

### Install browser-use with uv

Source: https://github.com/browser-use/browser-use/blob/main/skills/open-source/references/quickstart.md

Sets up a Python 3.12 virtual environment with uv, installs browser-use, and downloads Chromium via uvx.

```bash
pip install uv
uv venv --python 3.12
source .venv/bin/activate   # Windows: .venv\Scripts\activate
uv pip install browser-use
uvx browser-use install     # Downloads Chromium
```

--------------------------------

### SDK Setup for v3 (Python and TypeScript)

Source: https://github.com/browser-use/browser-use/blob/main/skills/cloud/references/api-v3.md

Setup for the v3 SDK in Python (async and sync) and TypeScript. The Python async client is recommended and uses the BROWSER_USE_API_KEY environment variable by default.

```python
# Python (async — recommended)
from browser_use_sdk.v3 import AsyncBrowserUse
client = AsyncBrowserUse()  # Uses BROWSER_USE_API_KEY env var

# Python (sync)
from browser_use_sdk.v3 import BrowserUse
client = BrowserUse()
```

```typescript
// TypeScript
import { BrowserUse } from "browser-use-sdk/v3";
const client = new BrowserUse();
```

--------------------------------

### Install OCI SDK

Source: https://github.com/browser-use/browser-use/blob/main/browser_use/llm/oci_raw/README.md

Install the OCI Python SDK, which is required for the ChatOCIRaw integration.

```bash
pip install oci
```

--------------------------------

### Start local MCP server with uvx

Source: https://github.com/browser-use/browser-use/blob/main/skills/cloud/references/guides/tools-integration.md

Run this command to start the local MCP server for MCP-native agents. The agent then discovers browser tools via the MCP protocol.

```bash
uvx --from 'browser-use[cli]' browser-use --mcp
```

--------------------------------

### Install and initialize OpenLIT

Source: https://github.com/browser-use/browser-use/blob/main/skills/open-source/references/monitoring.md

Zero-code OpenTelemetry instrumentation for browser-use. Install the openlit and browser-use packages, then call openlit.init() to auto-instrument.

```bash
pip install openlit browser-use
```

```python
import openlit

openlit.init()  # That's it — auto-instruments browser-use
```

--------------------------------

### Install browser-use-sdk with x402 extra

Source: https://github.com/browser-use/browser-use/blob/main/skills/x402/SKILL.md

Installs the SDK and x402 dependencies for Python (3.10+ required) or TypeScript. Choose based on the project language: package.json/tsconfig.json → TypeScript; pyproject.toml/requirements.txt/*.py → Python.

```bash
# Python (needs 3.10+ for the x402 extra)
pip install "browser-use-sdk[x402]"
# TypeScript
npm install browser-use-sdk @x402/fetch @x402/evm viem
```

--------------------------------

### Install Browser-Use with uv and activate environment

Source: https://github.com/browser-use/browser-use/blob/main/AGENTS.md

These commands create a Python 3.12 virtual environment with uv, activate it, and install browser-use along with Chromium. Use the Windows activation command on Windows.

```bash
pip install uv
uv venv --python 3.12
```

```bash
source .venv/bin/activate
# On Windows use `.venv\Scripts\activate`
```

```bash
uv pip install browser-use
uvx browser-use install
```

--------------------------------

### Browser-Use Cloud: Simple and Advanced Setup

Source: https://github.com/browser-use/browser-use/blob/main/skills/open-source/references/browser.md

Use the Browser-Use Cloud service by setting use_cloud=True. The advanced setup allows specifying a cloud profile ID, proxy country code, and timeout. Requires the BROWSER_USE_API_KEY environment variable.

```python
# Simple
browser = Browser(use_cloud=True)

# Advanced — bypasses captchas, geo-restrictions
browser = Browser(
    cloud_profile_id='your-profile-id',
    cloud_proxy_country_code='us',  # us, uk, fr, it, jp, au, de, fi, ca, in
    cloud_timeout=30,               # minutes (free: 15, paid: 240)
)
```

--------------------------------

### Install and initialize Laminar

Source: https://github.com/browser-use/browser-use/blob/main/skills/open-source/references/monitoring.md

Setup for Laminar monitoring integration. Install the lmnr package and initialize it, requiring the LMNR_PROJECT_API_KEY environment variable.

```bash
pip install lmnr
```

```python
from lmnr import Laminar

Laminar.initialize()  # Set LMNR_PROJECT_API_KEY env var
```

--------------------------------

### Initialize AsyncBrowserUse client (SETUP = B)

Source: https://github.com/browser-use/browser-use/blob/main/skills/x402/SKILL.md

Use this client initialization when SETUP = B. It reads the key and endpoint from your .env file. Mode A is accountless; Mode B requires an API key. Note that x402 requires the async client; the sync BrowserUse won't run when the x402 env var is set.

```python
from browser_use_sdk.v3 import AsyncBrowserUse

client = AsyncBrowserUse()                                  # Mode A (accountless)
# client = AsyncBrowserUse(api_key="bu_...")                # Mode B (top up existing account)
```

--------------------------------

### Setup virtual environment with uv

Source: https://github.com/browser-use/browser-use/blob/main/CLAUDE.md

Sets up a virtual environment with Python 3.11, activates it, and syncs dependencies using uv.

```bash
uv venv --python 3.11
source .venv/bin/activate
uv sync
```

--------------------------------

### Browser Methods: start, new_page, get_pages, get_current_page, close_page, stop

Source: https://github.com/browser-use/browser-use/blob/main/skills/open-source/references/actor.md

Core browser lifecycle and page management methods. Use to start the browser, open new tabs, list pages, get the active page, close tabs, and stop the browser.

```python
browser = Browser()
await browser.start()

page = await browser.new_page("https://example.com")  # Open new tab
pages = await browser.get_pages()                       # List all pages
current = await browser.get_current_page()              # Active page
await browser.close_page(page)                          # Close tab
await browser.stop()                                    # Cleanup
```

--------------------------------

### Create and run a Browser Use agent

Source: https://github.com/browser-use/browser-use/blob/main/AGENTS.md

Basic agent setup with a task and LLM, then running it asynchronously with a maximum step limit.

```python
from browser_use import Agent, ChatBrowserUse

agent = Agent(
    task="Search for latest news about AI",
    llm=ChatBrowserUse(),
)

async def main():
    history = await agent.run(max_steps=100)
```

--------------------------------

### Configure Oracle OCI with ChatOCIRaw

Source: https://github.com/browser-use/browser-use/blob/main/skills/open-source/references/models.md

Uses Oracle Cloud Infrastructure generative AI. Requires `~/.oci/config` setup and `pip install "browser-use[oci]"`. Auth types include `API_KEY`, `INSTANCE_PRINCIPAL`, and `RESOURCE_PRINCIPAL`.

```python
from browser_use import Agent, ChatOCIRaw

llm = ChatOCIRaw(
    model="meta.llama-3.1-70b-instruct",
    service_endpoint="https://inference.generativeai.us-chicago-1.oci.oraclecloud.com",
    compartment_id="your-compartment-id",
)
```

--------------------------------

### SDK Approach: Create and Stop Browser

Source: https://github.com/browser-use/browser-use/blob/main/skills/cloud/references/browser-api.md

Python SDK example to create a browser session, print the CDP and live URLs, and stop the session with refund of unused time.

```python
# Create browser
browser = await client.browsers.create(
    profile_id="uuid",
    proxy_country_code="us",
    timeout=60,
)

print(browser.cdp_url)   # wss://... for CDP connection
print(browser.live_url)  # View in browser

# Stop (unused time refunded)
await client.browsers.stop(browser.id)
```

--------------------------------

### GET /sessions/{session_id}

Source: https://github.com/browser-use/browser-use/blob/main/CLOUD.md

Retrieves detailed information about a specific session, including its status, live URL, start and end times, associated tasks, and public share URL. Requires the session ID and an API key.

```APIDOC
## GET /sessions/{session_id}

### Description
Get detailed session information including status, URLs, and task details.

### Method
GET

### Endpoint
https://api.browser-use.com/api/v2/sessions/{session_id}

### Parameters
#### Path Parameters
- **session_id** (string, uuid) - Required - The unique identifier of the session.

#### Header Parameters
- **X-Browser-Use-API-Key** (string) - Required - API key for authentication.

### Response
#### Success Response (200)
- **id** (string, uuid) - Session ID.
- **status** (SessionStatus) - Session status (active or stopped).
- **liveUrl** (string, nullable) - Live URL of the session.
- **startedAt** (string, date-time) - Session start time.
- **finishedAt** (string, date-time, nullable) - Session end time.
- **tasks** (array of TaskItemView) - List of tasks in the session.
- **publicShareUrl** (string, nullable) - Public share URL.

#### Response Example
{
  "id": "3fa85f64-5717-4562-b3fc-2c963f66afa6",
  "status": "active",
  "liveUrl": null,
  "startedAt": "2024-01-01T00:00:00Z",
  "finishedAt": null,
  "tasks": [
    {
      "id": "3fa85f64-5717-4562-b3fc-2c963f66afa6",
      "sessionId": "3fa85f64-5717-4562-b3fc-2c963f66afa6",
      "llm": "string",
      "task": "string",
      "status": "started",
      "startedAt": "2024-01-01T00:00:00Z",
      "finishedAt": null,
      "metadata": {},
      "output": null,
      "browserUseVersion": null,
      "isSuccess": null
    }
  ],
  "publicShareUrl": null
}

#### Error Responses
- **404** - Session not found.
- **422** - Validation Error.
```

--------------------------------

### Clone Repository and Navigate to App Folder

Source: https://github.com/browser-use/browser-use/blob/main/examples/apps/news-use/README.md

Clone the browser-use repository and change into the news-use example directory.

```bash
git clone https://github.com/browser-use/browser-use.git
cd browser-use/examples/apps/news-use
```

--------------------------------

### SDK Setup with v3 and v2

Source: https://github.com/browser-use/browser-use/blob/main/skills/cloud/references/guides/chat-ui.md

Initializes both SDK versions: v3 for sessions/messages and v2 for profiles. Uses NEXT_PUBLIC_BROWSER_USE_API_KEY, which exposes the key to the browser; move SDK calls to server actions or API routes in production.

```typescript
// api.ts
import { BrowserUse as BrowserUseV3 } from "browser-use-sdk/v3";
import { BrowserUse as BrowserUseV2 } from "browser-use-sdk";

const apiKey = process.env.NEXT_PUBLIC_BROWSER_USE_API_KEY ?? "";
const v3 = new BrowserUseV3({ apiKey });
const v2 = new BrowserUseV2({ apiKey });
```

--------------------------------

### PZERO ChatOpenAI setup

Source: https://github.com/browser-use/browser-use/blob/main/skills/open-source/references/models.md

Configure ChatOpenAI for PZERO using the PZERO_API_KEY environment variable. Use the /v1 base URL, pass catalog model ids as-is (no openai/ prefix), and note that the default deepseek-v4-flash model is text-only — set use_vision=False on the agent unless you pick a vision-capable model. List available models with GET https://api.pzero.studio/v1/models (no auth required).

```python
import os

llm = ChatOpenAI(
    model="deepseek-v4-flash",
    base_url="https://api.pzero.studio/v1",
    api_key=os.environ["PZERO_API_KEY"],
)
```

--------------------------------

### Run as MCP server with uvx

Source: https://github.com/browser-use/browser-use/blob/main/CLAUDE.md

Starts the library as an MCP server for integration with Claude Desktop. Requires the CLI extra.

```bash
uvx browser-use[cli] --mcp
```

--------------------------------

### Install TypeScript SDK

Source: https://github.com/browser-use/browser-use/blob/main/skills/cloud/references/quickstart.md

Install the browser-use-sdk package for TypeScript/JavaScript.

```bash
npm install browser-use-sdk
```

--------------------------------

### Clone repo and enter ad-use app

Source: https://github.com/browser-use/browser-use/blob/main/examples/apps/ad-use/README.md

Clone the browser-use repository and navigate to the ad-use example app directory.

```bash
git clone https://github.com/browser-use/browser-use.git
cd browser-use/examples/apps/ad-use
```

--------------------------------

### Verify browser-harness installation

Source: https://github.com/browser-use/browser-use/blob/main/skills/qa/SKILL.md

Verifies that browser-harness is installed and on PATH before running any QA tests. The command prints 'browser-harness OK' if the tool is available; otherwise it fails and you must install it.

```bash
command -v browser-harness && browser-harness <<'PY'
print("browser-harness OK")
PY
```

--------------------------------

### Run browser task with AsyncBrowserUse (SETUP = A)

Source: https://github.com/browser-use/browser-use/blob/main/skills/x402/SKILL.md

Use this script to run a browser task through the SDK when SETUP = A. The SDK reads BROWSER_USE_X402_PRIVATE_KEY from the environment and finds the x402 endpoint automatically. Write it as a throwaway script in /tmp or as a background job for long tasks, run it, report the result, and delete it.

```python
import asyncio, os
from browser_use_sdk.v3 import AsyncBrowserUse

async def main():
    client = AsyncBrowserUse()  # Mode B: AsyncBrowserUse(api_key=os.environ["BROWSER_USE_API_KEY"])
    result = await client.run(task="<the task>")
    print(result.output)

asyncio.run(main())
```

--------------------------------

### Start remote daemon and open a tab

Source: https://github.com/browser-use/browser-use/blob/main/browser_use/skills/browser-use/SKILL.md

These commands start a remote daemon with a made-up name (r7k2) and then open a new tab in that browser. Use the same name for BU_NAME when starting the daemon and when using it.

```bash
browser-use <<'PY'
start_remote_daemon("r7k2")
PY
```

```bash
BU_NAME=r7k2 browser-use <<'PY'
new_tab("https://example.com")
print(page_info())
PY
```

--------------------------------

### Get session live URL

Source: https://github.com/browser-use/browser-use/blob/main/skills/cloud/references/quickstart.md

Fetch session details to get the liveUrl for watching the agent in real-time.

```bash
curl https://api.browser-use.com/api/v2/sessions/<sessionId> \
     -H "X-Browser-Use-API-Key: $BROWSER_USE_API_KEY"
```

--------------------------------

### POST /tasks

Source: https://github.com/browser-use/browser-use/blob/main/CLOUD.md

Creates a new task. You can either start a new task (auto creates a new simple session) or start a new task in an existing session (you can create a custom session before starting the task and reuse it for follow-up tasks).

```APIDOC
## POST /tasks

### Description
Creates a new task. You can either start a new task (auto creates a new simple session) or start a new task in an existing session (you can create a custom session before starting the task and reuse it for follow-up tasks).

### Method
POST

### Endpoint
https://api.browser-use.com/api/v2/tasks

### Parameters
#### Header Parameters
- **X-Browser-Use-API-Key** (string) - Required - API key for authentication.

#### Request Body
- **task** (string) - Required - The task description.
- **llm** (SupportedLLMs) - Optional - The LLM to use. Supported values: browser-use-llm, gpt-4.1, gpt-4.1-mini, o4-mini, o3, gemini-2.5-flash, gemini-2.5-pro, gemini-flash-latest, gemini-flash-lite-latest, gemini-3-flash-preview, gemini-3.1-flash-lite, claude-sonnet-4-20250514, gpt-4o, gpt-4o-mini, llama-4-maverick-17b-128e-instruct, claude-3-7-sonnet-20250219.
- **startUrl** (string or null) - Optional - The starting URL for the task.
- **maxSteps** (integer) - Optional - Maximum number of steps.
- **structuredOutput** (string or null) - Optional - Structured output specification.
- **sessionId** (uuid or null) - Optional - The session ID to reuse an existing session.
- **metadata** (object or null) - Optional - Additional metadata as key-value pairs.
- **secrets** (object or null) - Optional - Secrets as key-value pairs.
- **allowedDomains** (array of strings or null) - Optional - List of allowed domains.
- **opVaultId** (string or null) - Optional - Op vault ID.
- **highlightElements** (boolean) - Optional - Whether to highlight elements.
- **flashMode** (boolean) - Optional - Whether to use flash mode.
- **thinking** (boolean) - Optional - Whether to enable thinking.
- **vision** (boolean or string) - Optional - Vision setting. Can be boolean or the string "auto".
- **systemPromptExtension** (string) - Optional - Extension to the system prompt.

### Request Example
{
  "task": "string",
  "llm": "browser-use-llm",
  "startUrl": "string",
  "maxSteps": 0,
  "structuredOutput": "string",
  "sessionId": "00000000-0000-0000-0000-000000000000",
  "metadata": {},
  "secrets": {},
  "allowedDomains": ["string"],
  "opVaultId": "string",
  "highlightElements": true,
  "flashMode": true,
  "thinking": true,
  "vision": true,
  "systemPromptExtension": "string"
}

### Response
#### Success Response (202)
- **id** (uuid) - The ID of the created task.
- **sessionId** (uuid) - The session ID associated with the task.

#### Response Example
{
  "id": "00000000-0000-0000-0000-000000000000",
  "sessionId": "00000000-0000-0000-0000-000000000000"
}

#### Error Responses
- **400** - Session is stopped or has running task.
- **404** - Session not found.
- **422** - Request validation failed.
- **429** - Too many concurrent active sessions.

```

--------------------------------

### Browser Constructor

Source: https://github.com/browser-use/browser-use/blob/main/skills/open-source/references/browser.md

Documentation for initializing a Browser instance with various remote/cloud configurations.

```APIDOC
## Browser(use_cloud=True)

### Description
Initializes a browser instance using the Browser-Use Cloud service. This is the recommended way to use a remote browser.

### Method
Constructor

### Parameters
- **use_cloud** (bool) - Required - Set to True to use the cloud service.

### Request Example
```python
browser = Browser(use_cloud=True)
```

### Response
Returns a Browser instance configured for cloud use.
```

```APIDOC
## Browser(cloud_profile_id, cloud_proxy_country_code, cloud_timeout)

### Description
Initializes a browser instance with advanced cloud settings to bypass captchas and geo-restrictions.

### Method
Constructor

### Parameters
- **cloud_profile_id** (string) - Optional - The profile ID to use for the cloud browser.
- **cloud_proxy_country_code** (string) - Optional - Country code for the proxy (us, uk, fr, it, jp, au, de, fi, ca, in).
- **cloud_timeout** (integer) - Optional - Timeout in minutes (free: 15, paid: 240).

### Request Example
```python
browser = Browser(
    cloud_profile_id='your-profile-id',
    cloud_proxy_country_code='us',
    cloud_timeout=30,
)
```

### Response
Returns a Browser instance configured with the specified cloud settings.
```

```APIDOC
## Browser(cdp_url)

### Description
Initializes a browser instance using a CDP (Chrome DevTools Protocol) URL from any provider.

### Method
Constructor

### Parameters
- **cdp_url** (string) - Required - The CDP endpoint URL of the remote browser.

### Request Example
```python
browser = Browser(cdp_url="http://remote-server:9222")
```

### Response
Returns a Browser instance connected to the specified CDP endpoint.
```

```APIDOC
## Browser(proxy, cdp_url)

### Description
Initializes a browser instance with proxy settings and an optional CDP URL.

### Method
Constructor

### Parameters
- **proxy** (ProxySettings) - Required - Proxy configuration object.
- **cdp_url** (string) - Optional - The CDP endpoint URL of the remote browser.

### Request Example
```python
from browser_use.browser import ProxySettings

browser = Browser(
    proxy=ProxySettings(
        server="http://proxy-server:8080",
        username="proxy-user",
        password="proxy-pass"
    ),
    cdp_url="http://remote-server:9222"
)
```

### Response
Returns a Browser instance configured with the specified proxy and CDP settings.
```

--------------------------------

### Basic Agent Usage with ChatBrowserUse

Source: https://github.com/browser-use/browser-use/blob/main/skills/open-source/references/agent.md

Basic setup for creating an agent with a task and LLM, then running it asynchronously with a maximum step limit. The default max_steps is 500, but it can be overridden in the run call.

```python
from browser_use import Agent, ChatBrowserUse

agent = Agent(
    task="Search for latest news about AI",
    llm=ChatBrowserUse(),
)

async def main():
    history = await agent.run(max_steps=500)
```

--------------------------------

### Webhook payload example

Source: https://github.com/browser-use/browser-use/blob/main/skills/cloud/references/features.md

Example webhook payload for the `agent.task.status_update` event. Contains task and session IDs, status, and metadata.

```json
{
  "type": "agent.task.status_update",
  "timestamp": "2025-01-15T10:30:00Z",
  "payload": {
    "task_id": "task_abc123",
    "session_id": "session_xyz",
    "status": "finished",
    "metadata": {}
  }
}
```

--------------------------------

### Get Profile OpenAPI Specification (GET /profiles/{profile_id})

Source: https://github.com/browser-use/browser-use/blob/main/CLOUD.md

OpenAPI 3.1.1 specification for the GET /profiles/{profile_id} endpoint. Requires the profile_id path parameter and the X-Browser-Use-API-Key header. Returns 200 with a ProfileView on success, 404 if the profile is not found, and 422 on validation error.

```yaml
openapi: 3.1.1
info:
  title: Get Profile
  version: endpoint_profiles.get_profile_profiles__profile_id__get
paths:
  /profiles/{profile_id}:
    get:
      operationId: get-profile-profiles-profile-id-get
      summary: Get Profile
      description: Get profile details.
      tags:
        - - subpackage_profiles
      parameters:
        - name: profile_id
          in: path
          required: true
          schema:
            type: string
            format: uuid
        - name: X-Browser-Use-API-Key
          in: header
          required: true
          schema:
            type: string
      responses:
        '200':
          description: Successful Response
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ProfileView'
        '404':
          description: Profile not found
          content: {}
        '422':
          description: Validation Error
          content: {}
components:
  schemas:
    ProfileView:
      type: object
      properties:
        id:
          type: string
          format: uuid
        name:
          type:
            - string
            - 'null'
        lastUsedAt:
          type:
            - string
            - 'null'
          format: date-time
        createdAt:
          type: string
          format: date-time
        updatedAt:
          type: string
          format: date-time
        cookieDomains:
          type:
            - array
            - 'null'
          items:
            type: string
      required:
        - id
        - createdAt
        - updatedAt
```

--------------------------------

### Browser()

Source: https://github.com/browser-use/browser-use/blob/main/AGENTS.md

Initializes a Browser instance with configurable options for headless mode, window size, and other display settings.

```APIDOC
## Browser Initialization

### Description
Creates a Browser instance with configurable options for headless mode, window size, and other display settings.

### Method
Constructor

### Endpoint
`Browser()`

### Parameters
#### Constructor Parameters
- **headless** (bool or None) - Optional - Run browser without UI. Auto-detects based on display availability.
- **window_size** (dict or ViewportSize) - Optional - Browser window size for headful mode. Use dict `{'width': 1920, 'height': 1080}` or `ViewportSize` object.
- **window_position** (dict, default: `{'width': 0, 'height': 0}`) - Optional - Window position from top-left corner in pixels.
- **viewport** (dict or ViewportSize) - Optional - Content area size, same format as `window_size`. Use `{'width': 1280, 'height': 720}` or `ViewportSize` object.
- **no_viewport** (bool or None) - Optional - Disable viewport emulation, content fits to window size.
- **device_scale_factor** (float) - Optional - Device scale factor (DPI). Set to `2.0` or `3.0` for high-resolution screenshots.
- **cdp_url** (str) - Optional - CDP URL for connecting to existing browser instance (e.g., `"http://localhost:9222"`).

### Request Example
```python
from browser_use import Browser

browser = Browser(
    headless=False,
    window_size={'width': 1000, 'height': 700},
)
```

### Response
#### Success Response
Returns a `Browser` instance that can be passed to an `Agent`.

### Notes
- The `Browser` instance also provides all Actor methods for direct browser control (page management, element interactions, etc.).
```

--------------------------------

### Configure environment variables

Source: https://github.com/browser-use/browser-use/blob/main/AGENTS.md

Copy the example environment file to .env and optionally set the logging level by uncommenting the BROWSER_USE_LOGGING_LEVEL line.

```bash
# Copy the example environment file
cp .env.example .env

# set logging level
# BROWSER_USE_LOGGING_LEVEL=debug
```

--------------------------------

### Example scheduler output

Source: https://github.com/browser-use/browser-use/blob/main/examples/apps/msg-use/README.md

Example JSON output from the scheduler, showing contact, original message, composed message, and scheduled time.

```json
[
  {
    "contact": "Magnus",
    "original_message": "Hi",
    "composed_message": "Hi",
    "scheduled_time": "2025-06-13 18:15"
  },
  {
    "contact": "Camila",
    "original_message": "I miss her",
    "composed_message": "I miss you ❤️",
    "scheduled_time": "2025-06-14 20:00"
  }
]
```

--------------------------------

### Initial login with QR code

Source: https://github.com/browser-use/browser-use/blob/main/examples/apps/msg-use/README.md

Run the one-time login script. Scan the QR code when the browser opens; the session is saved for future use.

```bash
python login.py
```