Playwright

Using Playwright MCP with PraisonAI

Playwright MCP

Playwright is a powerful browser automation tool that can be used with the Model Context Protocol (MCP) to enable agents to interact with web browsers.

Quick Start

python

from praisonaiagents import Agent, MCP

# Connect to Playwright MCP server (run: npx @playwright/mcp@latest --port 8931)
agent = Agent(
    instructions="You help search the web.",
    llm="gpt-4o-mini",
    tools=MCP("http://localhost:8931/sse")
)

agent.start("Find about Praison AI")

Installation

To use Playwright with MCP, you’ll need to install the Playwright MCP package:

bash

npm install @playwright/mcp

Running the Playwright MCP Server

You need to run the Playwright MCP server in a separate terminal before your agent can use it:

bash

npx @playwright/mcp@latest --port 8931

This will start a Playwright MCP server on port 8931 that your agent can connect to.

Using Playwright MCP with PraisonAI

Once the Playwright MCP server is running, you can create an agent that uses it as a tool:

python

from praisonaiagents import Agent, MCP

search_agent = Agent(
    instructions="""You help search the web.""",
    llm="gpt-4o-mini",
    tools=MCP("http://localhost:8931/sse")
)

search_agent.start("Find about Praison AI")

The agent will be able to use Playwright to automate browser interactions and search the web.

Advanced Configuration

You can customize the Playwright MCP server by passing additional options:

bash

npx @playwright/mcp@latest --port 8931 --browser chromium

For more information on available options, refer to the Playwright MCP documentation.