Summary

This skill provides comprehensive reference documentation for BigQuery's AI/ML capabilities, specifically `AI.FORECAST`, `AI.DETECT_ANOMALIES`, and `AI.GENERATE` functions. It teaches agents how to write SQL queries that leverage pre-trained TimesFM and Gemini models for time-series forecasting, anomaly detection, and generative AI tasks without requiring custom model training.

Detected Capabilities

Trigger Keywords

Phrases that MCP clients use to match this skill to user intent.

Referenced Domains

External domains referenced in skill content, detected by static analysis.

Use Cases

• Forecast future values in time series data using pre-trained models
• Detect anomalies and deviations in temporal datasets
• Generate text and structured content using generative AI within BigQuery
• Analyze multi-series time data with grouping columns
• Extract insights from historical data for anomaly detection baseline

Quality Notes

• Comprehensive syntax reference tables with all required and optional parameters clearly documented
• Practical, runnable SQL examples demonstrating each function with realistic use cases (Citibike dataset)
• Clear output schema definitions with column types and descriptions
• Well-organized reference structure with separate files for each function
• Good progressive complexity in examples (basic to multivariate scenarios)
• Missing: error handling guidance (e.g., minimum data point requirements mentioned but not contextualized in error scenarios)
• Missing: performance considerations or cost implications for running these functions
• Missing: limitations or data quality prerequisites beyond the '3 data points minimum' note
• Could benefit from troubleshooting section for common errors
• No guidance on authentication or connection setup for AI.GENERATE