markdown-html-orchestrator

Convert long markdown documents into single-file interactive HTML via deterministic routing.

25.3k|3.6k|Updated Oct 19, 2025
One-click install
npx skills add https://github.com/alirezarezvani/claude-skills --skill markdown-html-orchestrator
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: markdown-html-orchestrator
Source: https://github.com/alirezarezvani/claude-skills/tree/main/markdown-html/skills/markdown-html-orchestrator
Command: npx skills add https://github.com/alirezarezvani/claude-skills --skill markdown-html-orchestrator

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

Long markdown files (specs, RFCs, code reviews, slide decks) lose readability past roughly 100 lines, and manually converting them into shareable, navigable HTML is tedious and inconsistent.

Core Features & Use Cases

  • Deterministic doctype routing: A stdlib-only classifier scores filename and content signals to route input to one of three converter sub-skills (md-document, md-review, md-slides), refusing ambiguous or undersized inputs.
  • Hard pre-flight gates: Refuses conversion when input is under 100 lines, when the design-system onboarding has not been completed, or when the output directory is not writable.
  • Collision-safe output paths: Resolves output filenames with kebab slugs, doctype prefixes, and suffix/timestamp collision handling.
  • Use Case: Convert a 300-line PR writeup with diff blocks into a single self-contained HTML review page with severity-tagged annotations and jump navigation.

Quick Start

Ask the agent to convert your markdown file to HTML, for example: convert report.md into an interactive HTML document using the markdown-html orchestrator.

Frequently Asked Questions about markdown-html-orchestrator

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

FAQPage Schema
How do I convert a markdown file to a single HTML file?

Run the orchestrator on your markdown file; it classifies the document type and routes to the md-document, md-review, or md-slides converter, which writes one self-contained HTML file with inline CSS and JavaScript. The design-system onboarding must be completed first.

How does the markdown document type classifier work?

The doctype_classifier.py script scores filename hints (2 points each) and content signals like diff fences, slide boundaries, and table rows (1-2 points each). It routes silently when the winner scores 3 or more and doubles the runner-up, otherwise it asks one clarifying question.

Why does markdown to HTML conversion get refused for short files?

Conversion is refused when input is under 100 lines, based on the threshold that markdown remains the better format below that length. The classifier reports below_min_lines and the route explainer refuses before routing.

Can I convert markdown to HTML without running the design-system onboarding?

No. The orchestrator refuses if the design-system config at ~/.config/markdown-html/design-system.json is missing or lacks setup_completed_at. Run onboard.py once, or use --defaults for a zero-touch setup.

What external dependencies does the generated HTML use?

Only Google Fonts CSS and Prism.js from CDN are permitted; everything else is inlined into the single HTML file. Both degrade gracefully to system fonts and plain pre blocks if the CDN is blocked.

What happens when the output HTML file already exists?

The output path resolver suffixes the filename with -2, -3, and so on by default. You can opt into timestamped names or explicit overwrite with the --on-collision flag; silent overwrites never happen.