Add LangGraph agent example connecting to ToolHive vMCP

User Story

As a ToolHive user attempting to build my own agent service with LangGraph, I have no
reference implementation showing how to securely connect a LangGraph workflow to MCP
servers managed by ToolHive. Without this, I have to figure out the integration from
scratch.

Description

Create a self-contained LangGraph agent example in the toolhive repository that
demonstrates connecting a LangGraph graph/workflow to a ToolHive-managed vMCP endpoint
with Okta OIDC auth. LangGraph's graph-based architecture is popular for more complex
agentic workflows; this example should show how to wire MCP tool calls into a LangGraph
node.

Requirements

1. Create examples/agents/langgraph/ directory with a runnable Python example
2. Demonstrate how to initiate the SSO flow from the agent service to authenticate to a vMCP
3. Show the agent invoking at least one MCP tool through the vMCP via a LangGraph node
4. Include a README.md with prerequisites, setup steps, and how to run
5. Keep dependencies minimal and pinned

Acceptance Criteria

• Example is checked into examples/agents/langgraph/ in the toolhive repo
• A user can follow the README and run the example against their own vMCP
• The example demonstrates a LangGraph graph invoking at least one MCP tool
• The SSO flow is initiated from within the agent service
• README explains the LangGraph graph structure and how MCP fits in

Start with the new examples/agents/langgraph/ directory and its README.md, using the issue requirements as the implementation checklist. Build a runnable Python example that initiates SSO from the agent service, connects to a ToolHive-managed vMCP endpoint, and invokes an MCP tool through a LangGraph node. Done means the README covers prerequisites, setup, dependencies, graph structure, and running the example against a user's vMCP.