Python: FoundryAgent returns 400 if both agent_reference and tools are present in request body

[benke520]

Description

FoundryAgent fails with 400 when both agent_reference (agent specified) and tools are present in the request body.

RawFoundryAgentChatClient._prepare_options() validates that only FunctionTool objects are passed, but does not strip them from the serialized request. super()._prepare_options() includes them in the request body, and Foundry rejects tools when agent_reference is present.

Expected Behavior
_prepare_options() should strip tools from the outgoing request when agent_reference is present, while keeping them available in memory for FunctionInvocationLayer to execute locally.

Code Sample

from agent_framework import tool
from agent_framework.foundry import FoundryAgent
from azure.identity import AzureCliCredential

@tool(approval_mode="never_require")
async def get_weather(location: str) -> str:
    return f"Sunny in {location}"

agent = FoundryAgent(
    project_endpoint="https://...",
    agent_name="MyAgent",        # has get_weather defined server-side
    agent_version="1",
    credential=AzureCliCredential(),
    tools=[get_weather],         # local implementation
)

# This fails:
async for update in agent.run("What's the weather?", session=agent.create_session()):
    pass

Error Messages / Stack Traces

agent_framework.exceptions.ChatClientException: service failed to complete the prompt:
Error code: 400 - {'error': {'code': 'invalid_payload',
'message': 'Not allowed when agent is specified.', 'param': 'tools'}}

Package Versions

agent-framework: 1.0.0, agent-framework-foundry: 1.0.0

Python Version

3.10

Additional Context

OS: Windows

[mvanhorn]

Implemented in #5101 - strips tools, tool_choice, and parallel_tool_calls from run_options after agent_reference injection. FunctionTools still work client-side via the invocation layer.