by Andrea Silvestri and Jonathan Nyborg
March 20, 2026
Reviewing documentation to find the right WURFL capability is tedious—especially mid-implementation when you just need a quick answer. Today’s AI can generate Renaissance cat paintings, but the new WURFL Documentation MCP server puts AI to work for you: your AI assistant can answer WURFL questions directly, so you don’t have to dig through the docs.
What is MCP (Model Context Protocol)?
Before diving deeper into WURFL’s new capabilities, it’s helpful to understand the foundation: MCP.
LLMs once couldn’t tell you if a picnic was a good idea, since they lacked access to weather data. Now imagine if an LLM could connect to a forecast service before responding: “It’s going to rain—you should stay home.” MCP enables this. It enables LLMs to interact with legacy systems such as databases, APIs, and static files.
Anthropic introduced the Model Context Protocol just over a year ago, and it has rapidly become a popular standard. MCP enables LLMs to access external IT systems and documentation, allowing thousands of servers to deliver targeted, context-aware information for various applications. For developer documentation, this means you can use your AI assistant to directly query the source material—like the WURFL Documentation MCP server—and instantly receive ready-to-use code or guidance, instead of manually searching references.
Meet the WURFL Documentation MCP Server
The WURFL Documentation MCP Server connects your AI assistant directly to WURFL’s documentation, whether you’re working in Claude, VS Code Copilot, Cursor, or another tool. You might wonder: can’t you just search the docs yourself? Sure, but finding the right capability name is a bit like knowing a song exists but not being able to remember a single lyric to search for. You know what you need, you just can’t quite name it. With the MCP server configured, your AI assistant automatically calls the right documentation tools and returns a precise, implementation-ready answer, right inside your editor.
Get Started in Minutes
Assuming you are already using one of the popular AI Assistants, getting started with MCP is easier than you think, and yes, we’re aware of the irony of asking you to read setup documentation in an article about not having to read documentation. You can refer to our documentation:
If you have an account on the ScientiaMobile website, you already have the credentials to configure the WURFL Documentation MCP server. If not, go ahead and register: it’s free of charge. Once the server is configured, your AI assistant will automatically route WURFL-related questions to the right documentation tools with no manual searching required. Just ask:
”How do I query WURFL from Python to learn if a device is a tablet?”
Lean back and watch the magic happen.
Real-World Use Cases
WURFL covers 500+ device capability definitions. Here’s just a small selection of questions you can ask your AI assistant, which will call the WURFL MCP server’s documentation tools in real time to find the answer:
Device-Based Targeting
Developer Question:
“I need to target ads only to flagship smartphones. How do I filter for devices over $800?”
MCP Server Response:
Why it matters: A dashboard built on the wrong categorization logic will silently mislead every business decision that follows. The AI queries WURFL’s capability definitions in real time and steers you toward the right approach before you write a single line.
Identify Under-Served Devices
Developer Question:
“Our analytics show high bounce rates on certain devices. How do I identify which devices have poor experiences?”
MCP Server Response:
Why it matters: The AI connects to WURFL’s documentation tools to map your vague performance problem onto the specific capabilities that can quantify it, turning a frustrating mystery into a targeted, actionable query.
For another approach, ask the WURFL MCP server for implementation-ready code with the query: “How do I query the WURFL API from Python to learn if a device is a tablet?”
And that’s it. One question, working code. You’ll still need to fit it into your own structure, but the AI is happy to keep going, and follow-up questions are fair game.
The next level: Direct Device Intelligence
Think of the WURFL Documentation MCP Server as a very well-read colleague who has actually memorized the WURFL docs, one you can interrupt at any point to ask, “Wait, which capability does that again?” The next step is giving that colleague live access to the data itself.
You are probably wondering whether MCP can also be used to query WURFL directly by submitting a User-Agent or request headers and getting live device capability data back, right inside your AI assistant. That’s not what this MCP server does today, but it’s exactly where we’re headed. We’re actively working on it, and if you’d like early access, reach out to our sales team or write to our support team with your use case. We may be able to get you in ahead of the official release.
In the meantime, while we work on those advancements, you can start experiencing the benefits of AI-powered documentation today. Connect your AI assistant and get authoritative WURFL answers instantly.