code-viz-core

Generate self-contained HTML code walkthroughs from per-file JSON inputs.

Updated Apr 30, 2026
One-click install
npx skills add https://github.com/koseki2580/skills --skill code-viz-core
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: code-viz-core
Source: https://github.com/koseki2580/skills/tree/main/skills/code-viz-core
Command: npx skills add https://github.com/koseki2580/skills --skill code-viz-core

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

This Skill provides a safe, repeatable engine for turning multiple source files into a single, self-contained HTML “code walkthrough” without manually editing the HTML shell.

Core Features & Use Cases

  • Data-driven walkthrough rendering: author JSON data per file and render a side-by-side mentor rail walkthrough from it.
  • Scalable multi-file support: add more files by adding more data/*.json, then rebuild to generate one index.html.
  • Guardrails for truthfulness: includes an accuracy policy that distinguishes claims you can mark as certain vs inferred.
  • Clear boundaries between explanation and review: focuses on explanation/visualization while delegating deep review to review-oriented skills.
  • Use Case: generate an explain-code style interactive walkthrough for several functions in a module by producing one JSON payload per source file and compiling them into a single HTML artifact.

Quick Start

Use the code-viz-core skill to assemble a walkthrough by building the explainer directory agents/explainers/<name> into an interactive index.html, then open the produced index.html in a browser to verify the render.

Frequently Asked Questions about code-viz-core

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

FAQPage Schema
How do I generate a self-contained HTML code walkthrough from multiple source files?

To generate a self-contained HTML code walkthrough, you author per-file JSON inputs conforming to the walkthrough schema and run bundled build scripts to assemble them into a single static index.html file.

What is the best way to build side-by-side code visualizations for explain-code style content?

Building side-by-side code visualizations requires a data-driven rendering engine that compiles per-file JSON payloads into a static HTML shell, producing an interactive mentor rail walkthrough without manual HTML editing.

How do I scale a code walkthrough to support additional source files?

To scale a code walkthrough for more source files, you add a new data/*.json file for each source file and execute a rebuild, which automatically injects the updated CV_DATA and CV_META into the single output HTML.

Can I render git diffs and inline comments in an HTML code walkthrough?

Yes, the walkthrough rendering engine supports optional diff and comment rendering modes, allowing you to visualize code changes and annotations directly within the generated side-by-side HTML explainer.

How does the walkthrough generation process handle uncertain code explanations?

The generation process enforces an accuracy policy that requires claims to be explicitly marked as certain or inferred, ensuring truthfulness and clear boundaries between factual explanations and inferred logic.

When should I avoid using a single HTML file for code walkthrough generation?

You should avoid using single HTML walkthrough generation for deep architectural reviews, as this tool focuses strictly on explanation and visualization while delegating deep review tasks to review-oriented skills.