Use consistent terminology

utils/file_utils.py

@@ -23,7 +23,7 @@ import os
 from pathlib import Path
 from typing import Optional
 
-from .token_utils import MAX_CONTEXT_TOKENS, estimate_tokens
+from .token_utils import DEFAULT_CONTEXT_WINDOW, estimate_tokens
 
 logger = logging.getLogger(__name__)
 
@@ -508,14 +508,14 @@ def read_files(
     Args:
         file_paths: List of file or directory paths (absolute paths required)
         code: Optional direct code to include (prioritized over files)
-        max_tokens: Maximum tokens to use (defaults to MAX_CONTEXT_TOKENS)
+        max_tokens: Maximum tokens to use (defaults to DEFAULT_CONTEXT_WINDOW)
         reserve_tokens: Tokens to reserve for prompt and response (default 50K)
 
     Returns:
         str: All file contents formatted for AI consumption
     """
     if max_tokens is None:
-        max_tokens = MAX_CONTEXT_TOKENS
+        max_tokens = DEFAULT_CONTEXT_WINDOW
 
     logger.debug(f"[FILES] read_files called with {len(file_paths)} paths")
     logger.debug(

utils/model_context.py

@@ -72,7 +72,7 @@ class ModelContext:
         Returns:
             TokenAllocation with calculated budgets
         """
-        total_tokens = self.capabilities.max_tokens
+        total_tokens = self.capabilities.context_window
 
         # Dynamic allocation based on model capacity
         if total_tokens < 300_000:

utils/token_utils.py

@@ -9,7 +9,8 @@ approximate. For production systems requiring precise token counts,
 consider using the actual tokenizer for the specific model.
 """
 
-from config import MAX_CONTEXT_TOKENS
+# Default fallback for token limit (conservative estimate)
+DEFAULT_CONTEXT_WINDOW = 200_000  # Conservative fallback for unknown models
 
 
 def estimate_tokens(text: str) -> int:
@@ -32,9 +33,9 @@ def estimate_tokens(text: str) -> int:
     return len(text) // 4
 
 
-def check_token_limit(text: str) -> tuple[bool, int]:
+def check_token_limit(text: str, context_window: int = DEFAULT_CONTEXT_WINDOW) -> tuple[bool, int]:
     """
-    Check if text exceeds the maximum token limit for Gemini models.
+    Check if text exceeds the specified token limit.
 
     This function is used to validate that prepared prompts will fit
     within the model's context window, preventing API errors and ensuring
@@ -42,11 +43,12 @@ def check_token_limit(text: str) -> tuple[bool, int]:
 
     Args:
         text: The text to check
+        context_window: The model's context window size (defaults to conservative fallback)
 
     Returns:
         Tuple[bool, int]: (is_within_limit, estimated_tokens)
-        - is_within_limit: True if the text fits within MAX_CONTEXT_TOKENS
+        - is_within_limit: True if the text fits within context_window
         - estimated_tokens: The estimated token count
     """
     estimated = estimate_tokens(text)
-    return estimated <= MAX_CONTEXT_TOKENS, estimated
+    return estimated <= context_window, estimated