MCP (Model Context Protocol) server that returns Web-API Baseline status using Web Platform Dashboard API.
https://github.com/yamanoku/baseline-mcp-serverYou're deep in a conversation with Claude about implementing a new web feature when you hit the familiar wall: "Wait, is CSS Container Queries widely supported now?" Instead of breaking your flow to check Can I Use or MDN, what if your AI could instantly tell you the current Baseline status of any web API?
That's exactly what baseline-mcp-server delivers.
Every web developer knows this dance: you're architecting a solution, discussing trade-offs with an AI assistant, then suddenly need to verify browser support. You tab over to Can I Use, check a few browsers, maybe cross-reference MDN, then return to your conversation having lost your train of thought.
This MCP server eliminates that context switch by connecting your AI directly to the Web Platform Dashboard API. Now Claude, GPT, or any MCP-compatible AI can give you instant, accurate browser support data without you leaving the conversation.
The server taps into the same data source that powers web platform standardization decisions. When you ask about a web API, you get:
widely, newly, limited, no_data) — the standardized way to understand browser supportInstead of generic advice, your AI can now say: "CSS Grid is Baseline widely supported since 2020, safe to use in production" or "View Transitions API is newly Baseline as of 2024 — great for progressive enhancement but have a fallback ready."
Feature Planning: "What's the current support for CSS Subgrid?" Get instant Baseline status to make informed architecture decisions.
Code Reviews: Discussing whether a teammate's use of a new API is production-ready? Check its Baseline status without leaving your review context.
Progressive Enhancement: Planning fallback strategies? Know exactly which features need polyfills and which are safe to use directly.
Browser Targeting: Working on a project with specific browser requirements? Filter out browsers you don't support to get cleaner compatibility data.
Built on Deno with minimal dependencies, the server requires only network access to api.webstatus.dev. Add it to Claude Desktop or VS Code with a simple JSON config:
{
"mcpServers": {
"baseline-mcp-server": {
"command": "deno",
"args": [
"run",
"--allow-net=api.webstatus.dev",
"jsr:@yamanoku/baseline-mcp-server"
]
}
}
}
No database setup, no complex authentication, no heavyweight infrastructure. The server stays lightweight and focused on one thing: delivering accurate web platform data to your AI conversations.
Manual browser support checking is slow and error-prone. You might check Chrome and Firefox but forget about Safari's quirks, or miss that a feature just achieved Baseline status last month. The Web Platform Dashboard provides the authoritative, constantly updated view of what's actually shipping in browsers.
More importantly, having this data available in your AI conversations means better decisions get made in real-time. No more "I think that's supported" or "let me check and get back to you" — just immediate, data-driven guidance on what you can safely use in your code.
This MCP server transforms browser compatibility from a research task into automatic context your AI assistant always has at hand. Your development conversations become more informed, your architecture decisions more confident, and your workflow stays unbroken.
Ready to stop guessing about browser support? Get baseline-mcp-server running and let your AI assistant become your instant web platform compatibility expert.