Test that flock_get_debug_metrics returns the nested structure

Test batch processing with multiple audio files.

Test that audio type requires transcription_model.

Test that Ollama provider throws error for audio transcription.

Test llm_complete with multiple images in batch processing.

Test llm_complete with both image and text context.

Dedicated Ollama test - ensures Ollama is always exercised (local/free provider).

Dedicated Ollama image test - ensures Ollama image support is always exercised.

Test llm_complete with both audio and text context columns.

Test llm_complete with audio transcription using OpenAI.

Test llm_complete with image data integration.

Test embedding generation with multiple concatenated fields.

Test a realistic use case: embedding documents for similarity search.

Test error handling for malformed input structures.

Test that Ollama provider throws error for audio transcription in llm_filter.

Test llm_filter with multiple images in batch processing.

Test llm_filter with both image and text context.

Test llm_filter with audio transcription using OpenAI. The audio content says: 'Flock transforms DuckDB into a hybrid database and a semantic AI

Test llm_filter with image data integration.

Test llm_filter without context_columns parameter.

Test that Ollama provider throws error for audio transcription in llm_first.

Test basic llm_first functionality without GROUP BY.

Test llm_first behavior with empty table.

Test error handling with empty prompt.

Test error handling with non-existent model.

Test error handling with missing required arguments.

Test llm_first with multiple images in batch processing.

Test llm_first with images and GROUP BY clause.

Test metrics for llm_first aggregate function

Test llm_first with multiple selection criteria.

Test llm_first with audio transcription using OpenAI.

Test llm_first with batch size configuration.

Test llm_first with GROUP BY clause.

Test llm_first with image data integration.

Test llm_first with custom model parameters.

Test llm_first with special characters and unicode.

Test that Ollama provider throws error for audio transcription in llm_last.

Test basic llm_last functionality without GROUP BY.

Test llm_last behavior with empty table.

Test error handling with empty prompt.

Test error handling with non-existent model.

Test error handling with missing required arguments.

Test llm_last with multiple images in batch processing.

Test llm_last with images and GROUP BY clause.

Test llm_last with multiple evaluation criteria.

Test llm_last with audio transcription using OpenAI. The audio content says: 'Flock transforms DuckDB into a hybrid database and a semantic AI en

Test llm_last with batch size configuration.

Test llm_last with GROUP BY clause.

Test llm_last with image data integration.

Test llm_last with custom model parameters.

Test llm_last with special characters and unicode.

Test that Ollama provider throws error for audio transcription in llm_reduce.

Test basic llm_reduce functionality without GROUP BY.

Test llm_reduce behavior with empty table.

Test error handling with empty prompt.

Test error handling with non-existent model.

Test error handling with missing required arguments.

Test llm_reduce with multiple images in batch processing.