fix(openrouter): support monitoring cache tokens / reasoning tokens on OpenRouter response

[kargnas]

Currently, you can't know how much was cache read or cache write.

1. I added two cache token details:

Usage object fields

The usage object in API responses includes detailed cache metrics in the prompt_tokens_details field:

{
  "usage": {
    "prompt_tokens": 10339,
    "completion_tokens": 60,
    "total_tokens": 10399,
    "prompt_tokens_details": {
      "cached_tokens": 10318,
      "cache_write_tokens": 0
    }
  }
}

Document: https://openrouter.ai/docs/guides/best-practices/prompt-caching#usage-object-fields

2. Additionally, I also added reasoning_tokens too (but it's not about cache)

• Document: https://openrouter.ai/docs/api/reference/overview