Webpage Text Extractor — AI skill for Claude Code

**🇪🇸 Spanish version available:** [README_ES.md](README_ES.md) — Versión en español para público de habla hispana.

webpage-text-extractor

**Extract all text from a folder — like a web page saved with `Ctrl + S` in Chrome — and concatenate it into a single `.txt` file ready to send to an AI agent.**

[](https://www.python.org/) [](LICENSE) [](#requirements)

---

Why this script exists?

When you ask an AI agent (ChatGPT, Claude, a local assistant, etc.) to analyze a web page, the most practical approach is often not to pass the URL but **the actual content of the page**: its HTML, styles, and structure.

The workflow is:

1. Open the page in Google Chrome.
2. Press Ctrl + S (on macOS ⌘ + S) and choose "Webpage, Complete".
3. Chrome creates a folder with the page's HTML and all its resources (CSS, JS, images).
4. Run this script on that folder.
5. Get a single .txt file with the readable content of all text files, clearly separated by headers.

That `.txt` file is much easier to handle: you attach it to your AI agent, paste it in a chat, or process it as you wish, without having to send dozens of loose files.

Features

• Recursive: traverses all subdirectories with os.walk.
• Text detection: attempts to read each file as UTF-8; binary files (images, fonts, videos) are listed by name without cluttering output.
• Clear separators: each text file is preceded by a header with its full path:

  --- Text file content: Blog-example/page.html ---

• Zero dependencies: only Python standard library. No pip install needed.
• Two versions:
  • extract_text.py — version with command-line arguments (recommended).
  • extract_text_simple.py — original minimal version, edit two variables and run.

Requiremen