gen-metrics

What problem does it solve?

Manually defining MetricFlow metrics requires deep knowledge of MetricFlow syntax, semantic model structure, and business logic mapping, which is time-consuming and error-prone for data teams. This Skill eliminates that manual work by converting plain language business descriptions or existing SQL queries into valid, production-ready MetricFlow definitions.

Core Features & Use Cases

• Duplicate Detection & Reuse: Scans existing metrics to avoid redundant work and reuse existing measures from semantic models.
• Flexible Input Handling: Supports both single metric generation from natural language and batch processing of multiple SQL queries or CSV files with query data.
• End-to-End Validation & Sync: Automatically creates or updates semantic models, validates all generated metrics, runs dry-run SQL checks, and syncs final definitions to the knowledge base.
• Use Case: A data analyst can paste existing revenue SQL queries and ask the Skill to generate corresponding MetricFlow metrics, which will be validated and synced without manual YAML editing.

Quick Start

Provide a natural language description of your desired business metric, such as "count of unique paying users per month", and the Skill will generate, validate, and sync the corresponding MetricFlow metric definition.