FLUIDFLUID
  • Introduction
  • Quickstart
  • Why FLUID
  • FAQ
  • What FLUID Is
  • Core Principles
  • Agentic-Native Layer
  • FLUID vs ODCS / ODPS
  • Anatomy
  • Cheatsheet
  • Full Specification
  • Versions
  • JSON Schema 0.7.5 ↗
  • Reference (HTML) ↗
Examples
How-to
What's New
Deck
GitHub
GitHub
  • Introduction
  • Quickstart
  • Why FLUID
  • FAQ
  • What FLUID Is
  • Core Principles
  • Agentic-Native Layer
  • FLUID vs ODCS / ODPS
  • Anatomy
  • Cheatsheet
  • Full Specification
  • Versions
  • JSON Schema 0.7.5 ↗
  • Reference (HTML) ↗
Examples
How-to
What's New
Deck
GitHub
GitHub
  • Concepts

    • What FLUID Is (and Is Not)
    • Core Principles
    • The Looming Crisis of Context
    • The Agentic-Native Layer
    • FLUID vs ODCS / ODPS
    • The Reference Compiler — forge-cli

Core Principles

Five principles shape every FLUID contract. Together they make governance automated, ownership federated, and execution someone else's job.

PrincipleDescription
Data as a ProductFLUID's core mental model shifts from "pipelines" to "products." A pipeline is an imperative process; a product is a versioned asset with a defined interface, quality guarantees, and a clear owner. FLUID files are the specification for these products.
Declarative, Not ImperativeYou define the desired end state of your data product — what it consumes, what it exposes, and the contract it must adhere to. You do not define the step-by-step "how." That is the job of a FLUID-compliant tool, which reads your definition and figures out the best way to implement it.
Contracts as CodeThe contract block is the heart of every data product. It embeds schema, quality rules, and privacy treatments directly into a version-controlled file. This makes governance an automated, proactive part of the development lifecycle — not a reactive, manual process.
Federated OwnershipFLUID is designed for a Data Mesh. .fluid.yml files are decentralized and co-located with the domain teams that own them. The standard's use of globally unique data product ids lets a central orchestrator or catalog discover these distributed files and weave them into a single unified data fabric.
Compliant EcosystemFLUID is not a monolithic platform. It is a standard that delegates execution to the tools you already use. An orchestrator, a catalog, or an ingestion service becomes FLUID-aware by learning to read .fluid.yml files to configure itself. This fosters an open, composable ecosystem rather than creating a new silo.

Where to go next

  • What FLUID Is (and Is Not) — the conceptual anchor and the F.L.U.I.D philosophy.
  • The Agentic-Native Layer — how these principles play out for AI agents.
Edit this page on GitHub
Last Updated: 5/29/26, 5:26 PM
Contributors: fas89, Claude Opus 4.8
Prev
What FLUID Is (and Is Not)
Next
The Looming Crisis of Context