OKF Workbench

Product brief

Product brief

Summary

OKF Workbench is a local-first VS Code-compatible extension for creating, maintaining, validating, and visually exploring Open Knowledge Format bundles.

The product is intentionally broader than a 3D graph viewer. A viewer only helps teams that already have a healthy bundle. OKF Workbench is intended to help a user move from an empty folder to an actively maintained knowledge graph without leaving the editor.

Target users

Primary users:

Secondary users:

User problem

OKF is deliberately a minimal file format. That makes it portable, but leaves common workflow needs to consumers and authoring tools:

Value proposition

Start an OKF bundle, keep it healthy, and understand its structure without leaving the editor.

The main differentiator is the short path from discovery to repair:

find an orphan or broken link
-> inspect its context
-> open the source Markdown
-> edit and save
-> see diagnostics and graph update

Positioning

OKF Workbench should be positioned as an editor-native workbench, not as:

Product principles

Local first

Parsing, validation, indexing, and visualization run locally. The MVP has no telemetry, hosted account, LLM request, or content upload.

Source is primary

Markdown remains the source of truth. The graph is a derived view and must never hide or replace the underlying files.

Permissive consumption

The extension must tolerate unknown types, extra frontmatter fields, missing optional metadata, and broken links in accordance with OKF v0.1.

Safe generation

Create-only generation applies without an extra confirmation only when every target is absent and the workspace adapter can enforce no-overwrite creation. Any proposal that may update or replace an existing file is previewable and requires explicit approval. Existing user content is never silently overwritten. Re-running a generator should be idempotent wherever practical.

Useful without AI

All core features are deterministic. AI integration consists only of portable instruction templates for external agents.

Success indicators

Initial validation targets:

These are product hypotheses until measured in prototypes or user tests.