Debug and automate Safari browser sessions programmatically. For developers building AI agents that need to inspect DOM, execute JavaScript, and control Safari for testing or web automation.
io.github.HayoDev/safari-devtools-mcp
Local install
STDIO
No auth required
How models use it and what it is built for.
Debug and automate Safari browser sessions programmatically. For developers building AI agents that need to inspect DOM, execute JavaScript, and control Safari for testing or web automation.
Local install — runs as a subprocess.
Where to find authoritative docs and source for Safari DevTools.
Paste any of these into Agent Studio after connecting Safari DevTools.
Common questions about connecting and running Safari DevTools.
What can I do with the Safari DevTools MCP server?
You can inspect the DOM, execute JavaScript in the browser context, take screenshots, and automate user interactions like clicks and form submissions. It's designed for AI agents that need to interact with Safari programmatically for testing or web automation.
How do I install and run the Safari DevTools MCP server?
Install via npm with `npx safari-devtools-mcp@1.8.1`. The server uses stdio transport, so it communicates with your MCP client through standard input/output streams.
Does this work with headless Safari or only the GUI?
The registry metadata does not specify headless support. Check the project repository or documentation for details on whether it requires a running Safari instance with DevTools enabled.
What's the difference between this and using Puppeteer or Playwright?
This MCP server integrates Safari DevTools directly into an AI agent workflow via the Model Context Protocol, making it easier to use with Claude or other AI models. Puppeteer and Playwright are general-purpose browser automation libraries with broader browser support.
Are there any authentication or setup requirements?
The install command and registry data do not mention authentication. You'll need Safari running on your machine; consult the project repository for any additional setup steps or environment variables required.
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 StudioJudges Panel
45 judges that evaluate AI-generated code for security, cost, and quality with built-in AST.
Addozhang Nexus
Query Sonatype Nexus Repository (OSS/Pro) for Maven, Python, and Docker artifacts
Cyanheads MCP Ts Template
TypeScript template for building MCP servers with declarative tooling, observability, and auth.
Chernistry Bernstein
Declarative agent orchestration for engineering teams