Introduce strategic_token_limit to fix Anthropic bug

[winsonluk]

Fix for #1022

New config variable smart_token_limit defaults to 4000 to fix ValidationError thrown by ChatAnthropic when max_tokens=none, replacing it with max_tokens=4000.

[winsonluk]

@assafelovic here's the follow-up PR to #1022

INFO:     [07:18:11] 🤔 Planning the research strategy and subtasks (this may take a minute)...
Error with strategic LLM: 1 validation error for ChatAnthropic
max_tokens
  Input should be a valid integer [type=int_type, input_value=None, input_type=NoneType]
    For further information visit https://errors.pydantic.dev/2.9/v/int_type. Retrying with max_tokens=4000.
See https://github.com/assafelovic/gpt-researcher/issues/1022
Retrying with max_tokens=4000 successful.