Product Brief

Semantic Flow (sflo) - Product Brief

What is Semantic Flow?

Semantic Flow is a platform for creating, managing and publishing semantic meshes - dereferenceable, versioned collections of data resources where every IRI resolves to meaningful content.

Twin Purposes

  • Mint dereferenceable IRIs for referring to things on the Semantic Web
  • Hold versionable semantic data that uses those IRIs and can be referenced by other semantic data

Problems Solved

  • Free, Permanent, Self-Sovereign Data Storage - Provides individuals with free, permanent, self-describing data storage via the git provider of their choice
  • IRI Stability - Provides stable, dereferenceable IRIs for Semantic Web resources
  • Version Management - Tracks semantic data evolution with immutable version history
  • Content Dereferenceability - Every IRI resolves to meaningful HTML content
  • Transposability - Meshes can be moved between domains/projects without breaking internal links
  • Composability - Submeshes can be extracted and composed into larger structures

Components & Applications

sflo-host

The main host application that supports semantic mesh use and development. Built with:

  • Fastify web framework
  • Plugin architecture for extensibility
  • TypeScript for type safety

Plugins

Shared Packages

How It Works

Filesystem-Based Meshes

Meshes map directly from Git repository folder hierarchies to published static sites:

  • Every folder is a node (container for resources and child nodes)
  • Nodes contain components (flows, handles, assets, documentation)
  • Flows are versioned DatasetSeries (metadata, semantic data, arbitrary datasets, or config)
  • FlowShots are flow realizations (_default/, _working/, snapshot folders like 2025-11-24_0142_07_v1/)
  • Distributions are serialization files (TriG, JSON-LD, etc.)

The Weave Process

The weave process maintains mesh coherence and publication readiness:

  • Ensures required system components exist
  • Creates new snapshots from working data
  • Promotes working data to current
  • Updates metadata and provenance
  • Regenerates resource pages
  • Resolves internal links for transposability

User Experience Goals

For Developers

  • Git-native workflow - Meshes are just Git repositories
  • Static site deployment - Push to GitHub Pages or any static host
  • Type-safe development - TypeScript throughout
  • Plugin extensibility - Extend functionality through plugins

For Semantic Web Users

  • Dereferenceable IRIs - Every IRI resolves to content
  • Version history - Immutable snapshots for precise citation
  • Human-friendly - Resource pages provide context and navigation
  • Machine-readable - RDF distributions for automated processing

For AI Agents

  • Clear structure - Predictable folder/file organization

Backlinks