[Naveen Chatlapalli (1aifanatic)]

Motivation & Context

The TypeSafe MCP sample (packages/typesafe/samples/mcp_function_calling.py with mcp_weather_server.py) can't connect to its server under either supported mcp major version:

mcp installed	what fails
>=1.27.0,<2, as the README command and both PEP 723 headers pin	The Agent Framework MCP client: ToolException: Failed to create MCP session ... cannot import name 'Client' from 'mcp'. Since the MCP SDK 2.x migration, core's MCP tools need mcp 2.x; core's all extra pins mcp>=2.2.0,<3.
2.x, e.g. the workspace's 2.3.0	The server subprocess: No module named 'mcp.server.fastmcp' (FastMCP became MCPServer in 2.x). The client then reports Connection closed.

Description & Review Guide

• What are the major changes?
  • mcp_weather_server.py now uses mcp.server.mcpserver.MCPServer, the same API as samples/04-hosting/mcp/mcpserver_app.py. The tool passes structured_output=False so it still returns plain text (Seattle is sunny and 22 C.), matching the sample's documented output. Without it, mcp 2.x returns {"result": "..."} as the text content.
  • Both PEP 723 headers and the README run command now require mcp>=2.2.0,<3.
• What is the impact of these changes? Sample only.
• What do you want reviewers to focus on? Nothing special.

Verification: with the README's command (uv run --with "mcp>=2.2.0,<3" --package agent-framework-typesafe ...), MCPStdioTool launched the server, discovered get_weather, and returned Seattle is sunny and 22 C. for {"city": "Seattle", "detailed": false}. On main, both rows above fail. uv run --script mcp_weather_server.py also resolves and starts the server. I couldn't run the full sample because it needs a TYPESAFE_API_KEY. I found this by type-checking the example code shipped inside packages: pyright reported "FastMCP" is unknown import symbol.

Related Issue

N/A - trivial change: the sample is updated to the MCP SDK major version that Agent Framework's MCP client already requires.

AI Assistance

• No material AI assistance was used.
• This is an AI-assisted contribution. I reviewed, understood, and verified all submitted content and accept responsibility for it.

AI assistance details: Claude Code helped find this by type-checking the package samples, reproduce both failure modes, port the server, and draft this description.

Contribution Checklist

• The code builds clean without any errors or warnings
• All unit tests pass, and I have added new tests where possible
• The PR follows the Contribution Guidelines
• This PR links to an agreed issue with no competing open PR, or the Related Issue section documents a trivial-change or repository-automation exception.
• This is not a breaking change. If it is a breaking change, add the breaking change label (or add "[BREAKING]" to the title prefix, before or after any language prefix) — a workflow keeps the label and title prefix in sync automatically.

🤖 Generated with Claude Code

[Copilot]

🟢 Approval recommended

The narrow sample-only migration aligns with existing repository usage and dependency requirements, with no unresolved findings.

No open issues 🎉

What changed in this PR

Updates the TypeSafe MCP sample to MCP 2.x, aligning it with the Agent Framework MCP client’s dependency requirements.

Changes:

• Updates both script dependency headers and the README command to mcp>=2.2.0,<3.
• Replaces FastMCP with MCPServer.
• Disables structured output to preserve the documented plain-text weather response.

File	Description
python/​packages/​typesafe/​samples/​README.md	Updates the run command to require MCP 2.x.
python/​packages/​typesafe/​samples/​mcp_weather_server.py	Migrates the server API and dependency while preserving plain-text output.
python/​packages/​typesafe/​samples/​mcp_function_calling.py	Aligns the client script’s dependency with MCP 2.x.

_{Review effort: Balanced}