win4r

Okf Skill — Code Quality skill for Claude Code

Code Quality community

A stronger Claude Code skill/plugin for the Open Knowledge Format (OKF) v0.1: deterministic zero-dependency validator (hard errors vs lint), migrate/index/context/graph tools, examples, 137 tests, CI.

How to install Okf Skill

This entry records only its repository, not the path inside it, so there is no exact command to give. Open win4r/okf-skill and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Okf Skill does

Alternatives in Code Quality

Import Drawio — Redraw a draw.io file as an editorial diagram at a chosen format, size, and detail level 27.2k ★
Open Code Review — /open-code-review 21.5k ★
Job Stories — Create job stories with acceptance criteria in JTBD format 7.8k ★

README

OKF Skill — a stronger Claude Code skill set for the Open Knowledge Format

A Claude Code plugin for working with **[Open Knowledge Format (OKF)](https://github.com/GoogleCloudPlatform/knowledge-catalog/tree/main/okf)** v0.1 — Google Cloud's open spec for representing knowledge as a directory of Markdown files with YAML frontmatter (the "LLM-wiki" pattern). It helps you **create, maintain, validate, migrate, and visualize** OKF bundles.

The headline: the official OKF repo ships a *producer* (a Gemini/BigQuery enrichment agent) and an HTML *visualizer*, but **no standalone conformance validator** — the closest thing is a producer write-guard that is actually *stricter* than the spec. This project fills that gap with a **deterministic, zero-dependency validator** that implements the spec's conformance rules exactly and keeps **hard OKF errors strictly separate from optional lint warnings**.

No Gemini, Google Cloud, BigQuery, paid service, or account required. Pure Python 3.8+ standard library. Plain Markdown, YAML frontmatter, standard Markdown links, Git-friendly.

Highlights

Deterministic, zero-dependency validator (pure Python 3.8+ standard library) with a strict two-tier model: hard OKF v0.1 conformance errors are kept separate from optional lint warnings, plus an opt-in project-quality tier.
Bias toward acceptance. A zero-dependency YAML mini-parser whose accept/reject verdict is tested to match PyYAML on real bundles, so it never over-rejects spec-conformant content (multi-line folded scalars, scalar tags, relative links, unknown type values).
Full toolchain behind one okf.py CLI: scaffold (new), maintain indexes (index), progressive-disclosure context packs (context), an offline link-graph visualizer (graph, no CDN/backend), and Markdown/wiki → OKF migration with [[wikilink]] resolution (migrate).
A real skill set: an okf SKILL.md, slash commands, and reference docs, packaged as an