feat(context): enhance performance, caching, and settings management

- Replace hard-coded limits with configurable settings (e.g., cache memory size, truncation strategy, relevance settings).
- Optimize parallel execution in token counting, scoring, and reranking for source diversity.
- Improve caching logic:
  - Add per-context locks for safe parallel scoring.
  - Reuse precomputed fingerprints for cache efficiency.
- Make truncation, scoring, and ranker behaviors fully configurable via settings.
- Add support for middle truncation, context hash-based hashing, and dynamic token limiting.
- Refactor methods for scalability and better error handling.

Tests: Updated all affected components with additional test cases.

backend/app/services/context/budget/calculator.py

@@ -237,7 +237,7 @@ class TokenCalculator:
         """
         Count tokens for multiple texts.
 
-        Efficient batch counting with caching.
+        Efficient batch counting with caching and parallel execution.
 
         Args:
             texts: List of texts to count
@@ -246,13 +246,14 @@ class TokenCalculator:
         Returns:
             List of token counts (same order as input)
         """
-        results: list[int] = []
+        import asyncio
 
-        for text in texts:
-            count = await self.count_tokens(text, model)
-            results.append(count)
+        if not texts:
+            return []
 
-        return results
+        # Execute all token counts in parallel for better performance
+        tasks = [self.count_tokens(text, model) for text in texts]
+        return await asyncio.gather(*tasks)
 
     def clear_cache(self) -> None:
         """Clear the token count cache."""