WIP

- OpenRouter model configuration registry
- Model definition file for users to be able to control
- Update instructions

config.py

@@ -59,6 +59,12 @@ MODEL_CAPABILITIES_DESC = {
     "gemini-2.5-pro-preview-06-05": "Deep reasoning + thinking mode (1M context) - Complex problems, architecture, deep analysis",
 }
 
+# Note: When only OpenRouter is configured, these model aliases automatically map to equivalent models:
+# - "flash" → "google/gemini-flash-1.5-8b" 
+# - "pro" → "google/gemini-pro-1.5"
+# - "o3" → "openai/gpt-4o"
+# - "o3-mini" → "openai/gpt-4o-mini"
+
 # Token allocation for Gemini Pro (1M total capacity)
 # MAX_CONTEXT_TOKENS: Total model capacity
 # MAX_CONTENT_TOKENS: Available for prompts, conversation history, and files