```json { "queries": [ { "query": "SELECT model_name, SUM(calls) AS total_calls, SUM(input_tokens) AS total_input, SUM(output_tokens) AS total_output, SUM(cost_usd) AS total_cost FROM analysis_run_llm_usage GROUP BY 1 ORDER BY 5 DESC LIMIT 10", "description": "Total LLM usage and cost by model across all analysis runs." }, { "query": "SELECT provider, model, COUNT(*) AS call_count, AVG(latency_ms) AS avg_latency, SUM(cost_usd) AS total_cost FROM llm_calls WHERE status = 'ok' GROUP BY 1, 2 ORDER BY 5 DESC LIMIT 10", "description": "Successful LLM calls by provider and model with average latency and total cost." }, { "query": "SELECT purpose, COUNT(*) AS calls, AVG(input_tokens) AS avg_input, AVG(output_tokens) AS avg_output FROM llm_calls WHERE status = 'ok' GROUP BY 1 ORDER BY 2 DESC LIMIT 10", "description": "Most common LLM call purposes with average token usage." }, { "query": "SELECT dataset_id, name, status, table_count, total_rows FROM deleted_datasets WHERE deleted_at IS NOT NULL ORDER BY deleted_at DESC LIMIT 10", "description": "Recently deleted datasets with row and table counts." }, { "query": "SELECT type, status, COUNT(*) AS job_count, AVG(elapsed_ms) AS avg_duration FROM jobs GROUP BY 1, 2 ORDER BY 3 DESC", "description": "Job counts and average duration by type and status." }, { "query": "SELECT dataset_id, SUM(bytes_scanned) AS total_bytes, COUNT(*) AS probe_count, AVG(duration_ms) AS avg_duration FROM neal_probe_runs WHERE status = 'succeeded' GROUP BY 1 ORDER BY 2 DESC LIMIT 10", "description": "Datasets with highest data scanned during successful Neal probe runs." }, { "query": "SELECT kind, severity, COUNT(*) AS issue_count FROM quality_issues GROUP BY 1, 2 ORDER BY 3 DESC", "description": "Quality issues grouped by type and severity." }, { "query": "SELECT error_type, COUNT(*) AS error_count, AVG(cost_usd) AS avg_cost FROM llm_calls WHERE status = 'error' AND error_type IS NOT NULL GROUP BY 1 ORDER BY 2 DESC LIMIT 10", "description": "Most common LLM error types with average cost per failed call." }, { "query": "SELECT dataset_id, overall_rag, overall_score, data_completeness, referential_integrity FROM quality_scorecards WHERE overall_rag = 'green' ORDER BY overall_score DESC LIMIT 10", "description": "Top-quality datasets with green ratings and highest overall scores." }, { "query": "SELECT probe_kind, status, COUNT(*) AS run_count, AVG(finding_count) AS avg_findings FROM neal_probe_runs GROUP BY 1, 2 ORDER BY 3 DESC LIMIT 10", "description": "Neal probe run counts by type and status with average findings." } ] } ```