Introduction

docvera turns Git-managed Markdown into a validated, typed, queryable knowledge base. Your documents remain plain files. The engine checks the contracts around them.

Keep the files. Add confidence.

Markdown is easy to write and review. But a missing decision section, an invalid status, or a broken document link can quietly turn shared knowledge into guesswork.

With docvera you can:

• Validate metadata with JSON Schema, including required fields and allowed values.
• Check document structure with explicit heading and section rules.
• Connect documents through unique identities, typed metadata relations, Markdown links and anchors.
• Query knowledge with read-only SQL over documents, sections and diagnostics.
• Enforce cross-document policies with SHACL in the CLI and native Node.js engine.

Three pieces, one workspace

A workspace brings together Markdown documents, YAML or JSON schemas, and a docvera.yaml or docvera.toml configuration. The configuration maps document paths to their schemas.

knowledge/
├── docvera.yaml
├── schemas/
│   └── note.yaml
└── docs/
    └── note.md

There is no database server to provision and no derived index to keep in sync. CLI commands read the current files; the browser API receives an in-memory set of sources.

Choose your entry point

• CLI quickstart — initialize a workspace and use validation as a CI gate.
• Document schemas — describe metadata and required sections.
• SQL queries — explore your workspace as structured data.
• Browser & WASM — build an interactive integration.

Try it before installing

The playground executes the actual Rust engine compiled to WebAssembly. Start with a valid note, then introduce an invalid status to see the diagnostic.

Current boundaries

Docvera is a prototype. The browser backend supports portable validation and read-only GlueSQL, but not SHACL/RDF or filesystem loading. Native Node.js and the CLI expose additional operations. See the repository architecture notes for the full limitations.