Httrack Cloner Skill

httrack-cloner-skill is a self-contained workflow for reconstructing authorized websites as local React + TypeScript + TanStack Router projects. It captures deployed files with HTTrack, recovers runtime assets, removes automatic external requests, builds the result, verifies it in a browser, and opens the local preview.

[!IMPORTANT] Use this project only for websites and assets you are authorized to capture, modify, and redistribute.

Showcase

Captured website source and original third-party assets are not distributed in this repository.

What it does

Quick start

Install with your Agent

Copy this sentence into Codex, Claude Code, Cursor, or another coding agent:

Install httrack-cloner-skill for me from https://github.com/Redux0223/httrack-cloner-skill: first try `npx skills add Redux0223/httrack-cloner-skill -g`, then follow the repository requirements to install its dependencies, run the validation, and confirm that $httrack-cloner-skill is ready to use; if the installer is unavailable, use the manual installation steps in the README for my coding agent.

Or install it manually:

git clone https://github.com/Redux0223/httrack-cloner-skill.git
cd httrack-cloner-skill
npm ci --prefix scripts

mkdir -p ~/.codex/skills
ln -s "$(pwd)" ~/.codex/skills/httrack-cloner-skill

Give your coding agent a URL:

Use $httrack-cloner-skill on https://example.com/

Or run the deterministic entry point:

node scripts/run-url.mjs --url "https://example.com/" --authorized --depth 3

Resume a repair loop with:

node scripts/run-url.mjs --resume "/absolute/path/to/run"

Requirements

• Node.js 20 or newer and npm
• HTTrack available on PATH
• Chromium dependencies for Playwright
• Network access to the authorized source during capture

# macOS
brew install httrack

# Ubuntu / Debian
sudo apt-get update && sudo apt-get install -y httrack

npx --prefix scripts playwright install chromium

Output

clone-runs/<host>-<timestamp>-<id>/
├── mirror/       # captured source and inventory
├── react/        # editable React deliverable
├── proof/        # browser evidence
└── .cloner/      # trace, state, and repair history

Complex production bundles may remain behind a sanitized local React adapter when an automatic idiomatic rewrite would change behavior. The generated reports distinguish project delivery from visual parity instead of claiming pixel-perfect output without evidence.

Safety and limits

• Capture is host-scoped and requires --authorized.
• Automatic external runtime requests are blocked or reported.
• Live APIs require deterministic local replacements.
• Minified WebGL and canvas runtimes may need a local adapter.
• Authentication, CAPTCHA, DRM, paywalls, and access controls are out of scope.
• Generated authorization reports are technical evidence, not legal advice.

Development

npm ci --prefix scripts
npm run validate

The suite contains 123 automated tests covering capture, assets, React conversion, TanStack routing, external request cleanup, browser proof, repair loops, and reproducible delivery.

Read CONTRIBUTING.md before submitting changes. Report security issues through SECURITY.md.

License

The Skill implementation is released under the MIT License. Generated projects and captured assets retain their original ownership and licensing conditions.