refactor: implement _loop_response_model to manage response_model dur…

[lorenzejay]

…ing tool loops

This change introduces the _loop_response_model method in BaseAgentExecutor to conditionally return the response_model based on the presence of tools. The CrewAgentExecutor and AgentExecutor classes have been updated to utilize this new method, ensuring that the response_model is not sent to the LLM during tool loops, which prevents issues with structured-output schemas. Additionally, tests have been added to verify the correct behavior of this implementation.

---

Note

Medium Risk
Changes the parameters sent to LLM providers during tool-execution loops, which can affect structured-output behavior across multiple providers and execution paths. Test cassette updates reduce risk but regressions could still alter agent outputs or tool-calling flows.

Overview
Prevents response_model from being forwarded to the LLM when an executor has tools, via a new BaseAgentExecutor._loop_response_model() helper, to avoid providers returning schema-constrained placeholder JSON instead of making tool calls.

Updates all in-loop LLM call sites in CrewAgentExecutor (sync/async, ReAct and native tools) and experimental.AgentExecutor to route response_model through this helper, and adds regression tests enforcing the behavior.

Extends VCR header filtering (e.g., x-amz-security-token, x-crewai-organization-id) and refreshes Anthropic/Bedrock structured-output-with-tools cassettes and test models to match the new request/response flow.

^{Reviewed by Cursor Bugbot for commit e77d1c0. Bugbot is set up for automated code reviews on this repo. Configure here.}