[EroNiiChan]

Summary

Resolves #3760

Problem

In modelcontextprotocol/python-sdk, handler execution uniformly unpacked arguments as keyword arguments only (**kwargs):

• FuncMetadata.call_fn (src/mcp/server/mcpserver/utilities/func_metadata.py)
• Prompt.render (src/mcp/server/mcpserver/prompts/base.py)
• ResourceTemplate.create_resource (src/mcp/server/mcpserver/resources/templates.py)
• _resolve (src/mcp/server/mcpserver/resolve.py)

When handlers use positional-only arguments (def handler(x: str, /) or C-extension callables), invocation fails at runtime with:
TypeError: handler() got some positional-only arguments passed as keyword arguments: 'x'
or Pydantic ValidationError: Missing required positional only argument.

Solution (Minimal & Dependency-Free)

1. Added split_positional_args(fn, kwargs) helper to src/mcp/shared/_callable_inspection.py:
   • Inspects inspect.signature(fn) for inspect.Parameter.POSITIONAL_ONLY.
   • Returns ordered (tuple(pos_args), remaining_kwargs), preserving default values when optional parameters are not provided in kwargs.
   • Uses only Python standard library (inspect).
2. Integrated split_positional_args into the 4 handler invocation points so positional-only parameters are passed as *pos_args while remaining arguments are passed as **kwargs.
3. Added test coverage in tests/server/mcpserver/test_posonly_handlers.py verifying tool, prompt, and resource template execution with positional-only arguments.

[github-actions]

This PR has been closed automatically. This repo only keeps pull requests open when they come from a maintainer, or from a contributor a maintainer has assigned to the linked issue, and you aren't currently assigned to #3760.

If a maintainer assigns you to #3760, this PR reopens on its own and there's nothing more you need to do here. Assignment is a maintainer call based on capacity; comments that only ask to be assigned don't factor in. What does help is engaging on the issue itself by confirming the repro, explaining why it matters for your use case, or describing the approach you'd take.

You're welcome to keep pushing commits here (just avoid force-pushing, since GitHub can't reopen a rewritten branch), but that on its own won't get the PR reviewed or the issue assigned, and realistically most auto-closed PRs stay closed. There's no need to open a new PR either way.

CONTRIBUTING.md has the full reasoning, but in short:

• We're a small team with very little capacity to review community PRs right now.
• Many recent PRs are AI-generated with little human review, and reviewing one carefully still costs a maintainer as much time as it ever did. A well-described issue is usually more useful to us than the code.

Maintainers: reopen, remove missing-issue-link, or add bypass-issue-check to override.