Issue: Anthropic Messages API returns 400: tool_result.content.0.text.id from _convert_call_tool_result is rejected as “extra input”

## Problem When using `langchain-mcp-adapters` with Anthropic’s Messages API, the content produced by `_convert_call_tool_result` for a tool_result includes an `id` field under text. Anthropic’s schema rejects this and returns a 400 error. ``` messages.2.content.0.tool_result.content.0.text.id: Extra inputs are not permitted ``` This makes it impossible to use MCP tool results with Anthropic without post‑processing or overriding the adapter’s behavior. After [the recent breaking change](https://github.com/langchain-ai/langchain-mcp-adapters/pull/379), this issue started happening. ## Environment langchain-mcp-adapters: 0.2.0, 0.2.1 langchain-core: 1.2.2 anthropic Python SDK: 0.75.0 Python: 3.10 Anthropic model: claude-sonnet-4-5@20250929 ## Code Example repro_langchain_mcp_anthropic.py ```python import os import anthropic from langchain_mcp_adapters.tools import _convert_call_tool_result from mcp.types import CallToolResult, TextContent MODEL_NAME = os.environ.get("ANTHROPIC_MODEL", "claude-sonnet-4-5-20250929") def main() -> None: api_key = os.environ.get("ANTHROPIC_API_KEY") if not api_key: raise RuntimeError("Set ANTHROPIC_API_KEY in your environment before running.") client = anthropic.Anthropic(api_key=api_key) # Step 1: synthesize an MCP CallToolResult with simple text content. mcp_result = CallToolResult( content=[TextContent(type="text", text="test result")], structuredContent=None, isError=False, ) print("mcp_result:", mcp_result) # Step 2: pass it through the MCP adapter conversion. # This produces LangChain "standard" content blocks, which include an "id" field. tool_content, _artifact = _convert_call_tool_result(mcp_result) print("tool_content:", tool_content) # Step 3: Build Anthropic Messages API payload with tool_use + tool_result. # The tool_result.content is populated from the converted MCP content above. messages = [ { "role": "user", …