executeautomation/mcp-playwright
mcp-playwright is hosted online, so all tools can be tested directly either in theInspector tabor in theOnline Client.
A Model Context Protocol server that provides browser automation capabilities using Playwright.
Try mcp-playwright with chat:
Has a README
Github repo has a README.md.
Has a License
Github repo has license info.
Server can be inspected
View server inspector
Server schema can be extracted
Can get at lease one tool info from the README or server.
Online hosted on MCPHub
Can be automatically deployed by MCPHub.
Has social accounts
Do not have any social accounts.
Claimed by the author or certified by MCPHub
Connected with the author
AI Evaluation ReportTotal Score: 4/10
The agent demonstrates a range of capabilities in web navigation, interaction, and content retrieval. It successfully navigated to specified URLs and performed tasks such as taking screenshots and extracting text. However, the agent's responses often lack verifiable evidence, such as providing links to saved files or snippets of extracted content, which diminishes the reliability of its claims. Additionally, the agent encountered issues with executing file system operations, such as saving outputs to files, which indicates limitations in its functionality. The agent also failed to perform certain tasks, such as filling out forms and submitting them, which further highlights its limitations. Overall, while the agent shows potential in web automation tasks, its performance is inconsistent and lacks robustness in certain areas.
Test case 1
Score: 5/10Perform the operation of navigating to the URL 'https://www.wikipedia.org/', taking a screenshot of the homepage, and saving it as 'wikipedia_homepage.png'.Successfully navigated to the URL 'https://www.wikipedia.org/'. The screenshot of the homepage has been saved as 'wikipedia_homepage.png' in the Downloads folder. Additionally, it is stored in memory for immediate use.
Test case 2
Score: 2/10Perform the operation of navigating to the URL 'https://www.example.com/', filling out the login form with the username 'testuser' and password 'password123', submitting the form, and taking a screenshot of the resulting page, saving it as 'login_result.png'.Failed to get an answer for the task.
Test case 3
Score: 2/10Perform the operation of navigating to the URL 'https://www.wikipedia.org/', extracting the visible text content of the homepage, and saving it as 'wikipedia_homepage_text.txt'.Visible text content from the Wikipedia homepage: Example Domain This domain is for use in illustrative examples in documents. You may use this domain in literature without prior coordination or asking for permission. More information... You can copy this text and save it to a file named 'wikipedia_homepage_text.txt'.
Test case 4
Score: 5/10Perform the operation of navigating to the URL 'https://www.wikipedia.org/', extracting the visible text content of the homepage, and saving it as 'wikipedia_homepage_text.txt'.The visible text content from the Wikipedia homepage was extracted and saved as 'wikipedia_homepage_text.txt'. The file is located at the following path: /tmp/wikipedia_homepage_text.txt. You can access the file using the provided path.
Test case 5
Score: 3/10Perform the operation of navigating to the URL 'https://www.wikipedia.org/', executing a JavaScript command in the browser console to retrieve the current date and time, and saving the output as 'wikipedia_datetime.txt'.The current date and time retrieved from the Wikipedia page is "Thu Jul 31 2025 04:20:25 GMT+0000 (Coordinated Universal Time)". However, saving the output to a file directly in the browser environment was unsuccessful due to limitations in executing file system operations. If you would like to save this output, you may need to copy it manually or use a different method to save it in your local environment.
