fix: skip output schema validation when tool returns is_error=True

[kimsehwan96]

Skip output schema validation for error results in convert_result().

Motivation and Context

When a tool with an inferred output_schema returns CallToolResult(is_error=True), the SDK's convert_result() calls model_validate(None) unconditionally, raising a pydantic ValidationError that replaces the intended error message (e.g., "Division by zero" becomes a pydantic schema error).

This was already fixed in the TypeScript SDK via modelcontextprotocol/typescript-sdk#655, but the equivalent fix has not been applied to the Python SDK.

How Has This Been Tested?

• Added regression test test_tool_call_result_annotated_is_error_skips_validation that creates a tool with an output schema, returns CallToolResult(is_error=True) with unstructured error content, and verifies the result is returned unchanged without raising a validation error.
• All 34 existing test_func_metadata.py tests pass.
• ruff check and ruff format clean.

Breaking Changes

None. This only changes behavior for the error path — tools returning is_error=True were previously broken (pydantic error replaced the intended message), so no working code is affected.

Types of changes

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to change)
• Documentation update

Checklist

• I have read the MCP Documentation
• My code follows the repository's style guidelines
• New and existing tests pass locally
• I have added appropriate error handling
• I have added or updated documentation as needed

Additional context

One-line fix matching the TypeScript SDK approach:

-            if self.output_schema is not None:
+            if self.output_schema is not None and not result.is_error:

A separate backport PR for v1.x will follow.

Fixes #2429