What problem does it solve? Writing BigQuery scripts for Windmill requires knowing the correct parameter syntax, testing workflow, and metadata regeneration steps, and mistakes like deploying untested code or leaving stale lock files cause spurious diffs and broken runs. ## Core Features & Use Cases - BigQuery parameter syntax: Declares named script arguments using @name placeholders with typed comments, including defaults and S3Object file inputs decoded as JSON. - Correct test workflow: Distinguishes wmill script preview for local iteration from wmill script run for deployed versions, preventing accidental deploys of untested changes. - Metadata synchronization: Runs wmill generate-metadata after edits to keep .lock files, .script.yaml schemas, and wmill-lock.yaml hashes consistent. - S3 result streaming: Uses the -- s3 directive to stream large query results to S3 in JSON, Parquet, or CSV format, bypassing the 10000-row return cap. - Use Case: You need a Windmill script that queries a BigQuery users table with filterable parameters and exports results to S3 as Parquet; this Skill produces the correctly annotated SQL and previews it locally. ## Quick Start Write a Windmill BigQuery script that selects users by name and minimum age using @name parameters, then preview it locally with sample arguments.
