🎛 Playground

Pick a starter, edit the YAML, copy it. When you're ready, paste into a local file and run fluid validate.

In-browser validation via Pyodide is on the roadmap — for now the editor focuses on fast iteration on the schema without round-tripping through pip install.

What's in each template?

  • Local · DuckDB — runs on your laptop with platform: local + format: parquet. The fastest path from "never installed FLUID" to "deployed data product."
  • GCP · BigQuery — production-grade with schema, IAM grants in accessPolicy.grants[], AI/agent boundaries via agentPolicy, and column-level PII tagging.
  • AWS · Athena — S3-backed external table with the canonical bucket/prefix layout the AWS provider produces.
  • Snowflake — three-part-name binding with role-based access control.

All four are valid against fluid-schema-0.7.3, the current contract spec.

Next steps

After you've edited a contract you like:

# 1. Save the YAML to a local file (right-click → Save As, or just paste)
$ pbpaste > contract.fluid.yaml      # macOS — pulls from clipboard

# 2. Validate it (requires `pipx install data-product-forge` — see Getting Started)
$ fluid validate contract.fluid.yaml

# 3. Plan and apply against the local provider (no cloud account needed)
$ fluid plan contract.fluid.yaml
$ fluid apply contract.fluid.yaml --yes

Full quickstart → · CLI reference →