Add Wolfram MCP server integration

docs/integrations/wolfram.md

@@ -0,0 +1,100 @@
+---
+catalog_title: Wolfram
+catalog_description: Inject precise, real-time computation and knowledge
+catalog_icon: /integrations/assets/wolfram.svg
+catalog_tags: ["data", "mcp"]
+---
+
+# Wolfram MCP tool for ADK
+
+<div class="language-support-tag">
+  <span class="lst-supported">Supported in ADK</span><span class="lst-python">Python</span><span class="lst-typescript">TypeScript</span>
+</div>
+
+The [Wolfram MCP Server](https://www.wolfram.com/artificial-intelligence/mcp-service/)
+transforms your AI environment into a rigorous computational powerhouse. By
+integrating [Wolfram Language](https://www.wolfram.com/language/) and
+[Wolfram|Alpha](https://www.wolframalpha.com/), this server provides access to
+curated data and sophisticated algorithms. When a user submits a query, the LLM
+works with Wolfram to convert it into Wolfram Language for precise evaluation,
+and these exact results are incorporated into the response provided by the LLM.
+
+Whether you are solving complex differential equations, analyzing chemical
+structures or querying real-time socioeconomic data, the MCP Server ensures
+your AI has the "computational brain" necessary to deliver verified, high-level
+technical results that can be trusted.
+
+## Use cases
+
+- Solving complex differential equations
+- Analyzing chemical structures
+- Querying real-time socioeconomic data
+
+## Prerequisites
+
+- No special requirements
+- No auth needed
+
+## Use with agent
+
+=== "Python"
+
+    === "Remote MCP Server"
+
+        ```python
+        from google.adk.agents import Agent
+        from google.adk.tools.mcp_tool import McpToolset
+        from google.adk.tools.mcp_tool.mcp_session_manager import StreamableHTTPConnectionParams
+
+        root_agent = Agent(
+            model="gemini-flash-latest",
+            name="wolfram_agent",
+            instruction="Use Wolfram tools for computation and knowledge queries.",
+            tools=[
+                McpToolset(
+                    connection_params=StreamableHTTPConnectionParams(
+                        url="https://agenttools.wolfram.com/mcp",
+                    ),
+                )
+            ],
+        )
+        ```
+
+=== "TypeScript"
+
+    === "Remote MCP Server"
+
+        ```typescript
+        import { LlmAgent, MCPToolset } from "@google/adk";
+
+        const rootAgent = new LlmAgent({
+            model: "gemini-flash-latest",
+            name: "wolfram_agent",
+            instruction: "Use Wolfram tools for computation and knowledge queries.",
+            tools: [
+                new MCPToolset({
+                    type: "StreamableHTTPConnectionParams",
+                    url: "https://agenttools.wolfram.com/mcp",
+                }),
+            ],
+        });
+
+        export { rootAgent };
+        ```
+
+## Available tools
+
+Tool | Description
+---- | -----------
+`WolframContext` | Uses semantic search to retrieve any relevant information from Wolfram
+`WolframLanguageEvaluator` | Evaluates Wolfram Language code for the user in a Wolfram Language kernel
+`WolframAlpha` | Use natural language queries with Wolfram\|Alpha to get up-to-date computational results about entities in chemistry, physics, geography, history, art, astronomy, and more
+
+These are documented in
+[Wolfram/AgentTools/tutorial/DefaultTools](https://reference.wolfram.com/language/Wolfram/AgentTools/tutorial/DefaultTools.html).
+
+## Additional resources
+
+- [Wolfram MCP Service Documentation](https://www.wolfram.com/artificial-intelligence/mcp-service/)
+- [Wolfram AgentTools Repository](https://github.com/WolframResearch/AgentTools)
+- [Wolfram Support](https://support.wolfram.com/)