FIX: Adding openai invalid_prompt safety blocks as content filters

[rlundeen2]

When running against GPT-5.4, it raises this filter at the model level:

Root cause: BadRequestError: Error code: 400 - {'error': {'message': "Invalid prompt: we've limited access to this content for safety reasons. This type of information may be used to benefit or to harm people. We are continuously refining our work in this area, and you can read more about our approach in our blog post (https://openai.com/index/preparing-for-future-ai-capabilities-in-biology) and Model Spec (https://openai.com/index/introducing-the-model-spec).", 'type': 'invalid_request_error', 'param': None, 'code': 'invalid_prompt'}}

This PR handles this as a filter.