Overview

What is SchemaFlux?

SchemaFlux is a compiler for structured data. You define entities as markdown files with YAML frontmatter, declare a schema in a config file, and SchemaFlux handles the transformation into a complete static site.

markdown + frontmatter  ->  frontend  ->  12 passes  ->  backend  ->  output

Zero external dependencies. Single static binary. Go standard library only.

How It Works

SchemaFlux operates on entities -- units of structured data with fields, taxonomies, and relationships. A config file defines the schema; a pipeline of passes resolves slugs, sorts, enriches, groups, computes relationships, and validates. Backends consume the resulting intermediate representation (IR) to produce output.

1,997 entities -> 2,328 pages in ~500ms

The compiler pipeline has three stages:

1. Frontend parses markdown files with YAML frontmatter into an intermediate representation
2. Passes transform the IR: slug resolution, sorting, enrichment, taxonomy grouping, related entity scoring, graph enrichment, content analysis, URL resolution, schema generation, validation
3. Backend emits output from the finalized IR

The IR is immutable once passes complete -- backends read but never modify.

Key Features

• Zero dependencies -- only the Go standard library, no third-party packages
• Single binary -- compiles to a single static executable
• Pluggable backends -- the HTML backend ships built-in; others can be added
• 12-pass pipeline -- ordered, dependency-aware transformations
• Taxonomy system -- automatic grouping, pagination, A-Z indices, letter pages
• Structured data -- JSON-LD, Open Graph, sitemaps, RSS feeds, robots.txt
• Custom markdown renderer -- headings, lists, tables, code blocks, inline formatting
• Custom YAML parser -- zero-dependency frontmatter and config parsing
• Template engine -- Go html/template with 60+ helper functions
• Search index -- JSON search index generation for client-side search
• LLMs.txt -- automatic generation of llms.txt for AI crawlers
• Concurrent rendering -- 32-goroutine pool for entity page rendering

Architecture

compiler.Compile(cfg)
  -> frontend.Parse()        # markdown + YAML -> IR
  -> pass.Registry.RunAll()  # 12 ordered passes
  -> backend.Emit()          # IR -> output

The codebase is organized under internal/:

Use Case: Static Sites

The built-in HTML backend compiles structured data into a complete static site with taxonomy pages, pagination, A-Z indices, search index, JSON-LD, Open Graph, sitemaps, RSS, and llms.txt.