What is SHAFT Engine and how does it differ from raw Selenium?

SHAFT Engine is a Selenium-based framework that abstracts common web automation patterns and provides a higher-level API for test writing. It simplifies cross-browser testing and element interaction compared to writing raw Selenium code.

How do I install and run the SHAFT MCP server?

Use Docker to run the server: `docker run ghcr.io/shafthq/shaft-mcp:10.2.20260501`. The server communicates via stdio transport, so connect it to your MCP client (Claude, etc.) following your client's MCP configuration docs.

Can I run tests in parallel or on multiple browsers at once?

SHAFT Engine supports cross-browser testing configuration. Check the SHAFT documentation or ask the MCP server about parallel execution and browser capability setup for your specific test needs.

What are the system requirements and dependencies?

The Docker image includes all dependencies. You need Docker installed and a compatible host OS. The image is based on SHAFT Engine 10.2.20260501, which requires Java and browser drivers (handled by the container).

Is SHAFT Engine free and open-source?

SHAFT Engine is open-source. The MCP server is distributed via GitHub and Docker Hub. Check the ShaftHQ GitHub repository for licensing details and contribution guidelines.

MCP ServerSTDIOOfficialv10.2.20260501

SHAFT MCP Server

Automate web testing and browser interactions using SHAFT Engine, a Selenium-based framework. Built for QA engineers and developers who need reliable cross-browser test automation via an MCP server.

io.github.ShaftHQ/shaft-mcp

Try MCP with AI models View source

Hosted URL

Local install

Transport

STDIO

Auth

No auth required

What the SHAFT MCP server does

How models use it and what it is built for.

Automate web testing and browser interactions using SHAFT Engine, a Selenium-based framework. Built for QA engineers and developers who need reliable cross-browser test automation via an MCP server.

Connect to SHAFT

Local install — runs as a subprocess.

docker run ghcr.io/shafthq/shaft-mcp:10.2.20260501

Resources

Where to find authoritative docs and source for SHAFT.

Example prompts for SHAFT

Paste any of these into Agent Studio after connecting SHAFT.

Write a test script that logs into a web app and verifies the dashboard loads
How do I configure SHAFT Engine to run tests across Chrome, Firefox, and Safari
Generate a test that fills out a form, submits it, and validates the success message
What locator strategies does SHAFT Engine support for finding elements on a page

SHAFT MCP server — FAQ

Common questions about connecting and running SHAFT.

What is SHAFT Engine and how does it differ from raw Selenium?
SHAFT Engine is a Selenium-based framework that abstracts common web automation patterns and provides a higher-level API for test writing. It simplifies cross-browser testing and element interaction compared to writing raw Selenium code.
How do I install and run the SHAFT MCP server?
Use Docker to run the server: `docker run ghcr.io/shafthq/shaft-mcp:10.2.20260501`. The server communicates via stdio transport, so connect it to your MCP client (Claude, etc.) following your client's MCP configuration docs.
Can I run tests in parallel or on multiple browsers at once?
SHAFT Engine supports cross-browser testing configuration. Check the SHAFT documentation or ask the MCP server about parallel execution and browser capability setup for your specific test needs.
What are the system requirements and dependencies?
The Docker image includes all dependencies. You need Docker installed and a compatible host OS. The image is based on SHAFT Engine 10.2.20260501, which requires Java and browser drivers (handled by the container).
Is SHAFT Engine free and open-source?
SHAFT Engine is open-source. The MCP server is distributed via GitHub and Docker Hub. Check the ShaftHQ GitHub repository for licensing details and contribution guidelines.

Skip the local setup — run MCP in your browser

MCP Playground runs 10,000+ hosted MCP servers — GitHub, Linear, Notion, Stripe, Sentry and more — across Claude, GPT, Gemini, DeepSeek and 60+ AI models. Compare model answers side-by-side, save agent presets, share runs. Zero install.

Open Agent Studio