set temperature-0.0 to improve consistency [AUTH-1184] (#17096)

<!--
Thanks for taking the time to open a Pull Request (PR)! Please make sure
you've read the "Opening Pull Requests" section of our Contributing
Guide:


https://github.com/Opentrons/opentrons/blob/edge/CONTRIBUTING.md#opening-pull-requests

GitHub provides robust markdown to format your PR. Links, diagrams,
pictures, and videos along with text formatting make it possible to
create a rich and informative PR. For more information on GitHub
markdown, see:


https://docs.github.com/en/get-started/writing-on-github/getting-started-with-writing-and-formatting-on-github/basic-writing-and-formatting-syntax

To ensure your code is reviewed quickly and thoroughly, please fill out
the sections below to the best of your ability!
-->

# Overview
Set temperature parameter  to 0.0 to reduce hallucination.

Closes AUTH-1184
<!--
Describe your PR at a high level. State acceptance criteria and how this
PR fits into other work. Link issues, PRs, and other relevant resources.
-->

## Test Plan and Hands on Testing
CI handles testing.

<!--
Describe your testing of the PR. Emphasize testing not reflected in the
code. Attach protocols, logs, screenshots and any other assets that
support your testing.
-->

## Changelog
only temperature parameter
<!--
List changes introduced by this PR considering future developers and the
end user. Give careful thought and clear documentation to breaking
changes.
-->

## Review requests
Open UI and try to generate. All should be normal.

<!--
- What do you need from reviewers to feel confident this PR is ready to
merge?
- Ask questions.
-->

## Risk assessment
Low
<!--
- Indicate the level of attention this PR needs.
- Provide context to guide reviewers.
- Discuss trade-offs, coupling, and side effects.
- Look for the possibility, even if you think it's small, that your
change may affect some other part of the system.
- For instance, changing return tip behavior may also change the
behavior of labware calibration.
- How do your unit tests and on hands on testing mitigate this PR's
risks and the risk of future regressions?
- Especially in high risk PRs, explain how you know your testing is
enough.
-->

opentrons-ai-server/api/domain/anthropic_predict.py

@@ -95,6 +95,7 @@ def _process_message(
             tools=self.tools,  # type: ignore
             extra_headers={"anthropic-beta": "prompt-caching-2024-07-31"},
             metadata={"user_id": user_id},
+            temperature=0.0,
         )
 
         logger.info(