QDOC - Progressive Documentation

Generate and update project documentation following a three-tier progressive docs pattern.

Updated Jan 29, 2026
One-click install
npx skills add https://github.com/sparkst/sparkry-claude-skills --skill qdoc-progressive-documentation-sparkst
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: QDOC - Progressive Documentation
Source: https://github.com/sparkst/sparkry-claude-skills/tree/main/plugins/qshortcuts-support/skills/qdoc
Command: npx skills add https://github.com/sparkst/sparkry-claude-skills --skill qdoc-progressive-documentation-sparkst

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Keeping READMEs, component context files, and changelogs in sync with code changes is tedious and often skipped, leaving documentation stale and misleading. This Skill automates documentation updates after feature work by analyzing git diffs and requirement IDs. ## Core Features & Use Cases - Progressive Docs Generation: Produces a three-tier structure — root README (mental model), domain READMEs (boundaries and patterns), and .claude-context files (AI-assistant context). - CHANGELOG Management: Maintains Keep a Changelog format with semantic versioning and REQ ID traceability for every entry. - Scoped Updates: Supports depth and path flags to document a single domain, bump a version, or refresh the whole project. - Use Case: After implementing an OAuth feature tracked as REQ-42, run the skill to update the auth domain README, create a .claude-context file, and add a versioned CHANGELOG entry automatically. ## Quick Start Ask the AI to run QDOC to analyze recent git changes and update the READMEs, context files, and CHANGELOG accordingly.

Frequently Asked Questions about QDOC - Progressive Documentation

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I keep project READMEs in sync with code changes?▼

Run the documentation workflow after each feature implementation. It analyzes git diff output to identify changed files and affected domains, then updates the corresponding root README, domain READMEs, and .claude-context files automatically.

How to generate a CHANGELOG with semantic versioning?▼

Use the version flag such as --version=minor to cut a release entry. Entries follow the Keep a Changelog format with Added, Changed, Fixed, and Deprecated sections, each referencing requirement IDs for traceability.

What is the progressive documentation pattern?▼

It is a three-tier structure: a root README under 200 words giving a mental model, domain READMEs under 500 words covering boundaries and patterns, and .claude-context files under 300 words providing AI assistants with domain-specific context.

Can I document only one domain instead of the whole project?▼

Yes, pass a path and depth flag such as a domain directory with --depth=domain. This updates only that domain's README and .claude-context file, skipping the root README and CHANGELOG.

Why does my root README keep growing too long?▼

Root READMEs should stay under 200 words and link to domain READMEs for details. Move architecture and pattern details into the relevant domain README to restore the progressive disclosure structure.