Thank you for your contribution! I've reviewed the changes and this PR successfully fixes the Ollama max context window display issue. The implementation follows the same pattern as LM Studio, which provides good consistency.

✅ Positive Aspects

• Correctly addresses the root cause where info was always undefined
• Good consistency with LM Studio's implementation pattern
• Comprehensive test coverage added for the new functionality
• Type changes are properly propagated through the codebase

💭 Minor Suggestions for Improvement

1. Cache flushing in requestOllamaModels handler (webviewMessageHandler.ts line ~662)

   • The cache flush ensures fresh models, which is good for accuracy
   • However, if this handler is called frequently (e.g., on every settings page load), it might impact performance
   • Consider whether the cache flush is always necessary, or if it could be conditional

2. Error message clarity (useOllamaModels.ts line ~16)

   • The timeout error message could be more helpful
   • Consider: "Ollama models request timed out. Please check if Ollama is running and accessible at the configured URL."

3. Conditional fetching logic (useOllamaModels.ts line ~39)

   • The hook only fetches models when modelId is provided
   • This differs from the LM Studio implementation
   • While this seems intentional, a comment explaining the reasoning would help future maintainers

Overall Assessment

The implementation is solid and fixes the reported issue effectively. The code quality is good and follows existing patterns. The minor suggestions above are optional improvements that don't block the PR.

Thank you @ItsOnlyBinary!