refactor(provider): simplify LiteLLM requester usage handling

- Remove unused Anthropic-specific tool schema generation
  - Share completion argument construction between normal and streaming calls
  - Use LiteLLM/OpenAI native usage fields for monitoring
  - Collect stream token usage from LiteLLM stream_options
  - Update LiteLLM requester tests for unified usage fields

tests/unit_tests/provider/test_litellmchat.py

@@ -93,8 +93,6 @@ class TestExtractUsage:
         assert result['prompt_tokens'] == 100
         assert result['completion_tokens'] == 50
         assert result['total_tokens'] == 150
-        assert result['input_tokens'] == 100  # Compatibility alias
-        assert result['output_tokens'] == 50  # Compatibility alias
 
     def test_extract_usage_with_zero_values(self):
         """Test extraction when values are 0"""