Content and Archival Assistant Stripping Ads, Menus, and Surrounding Clutter from Webpages to Transfer Complete, Unsplit Text Verbatim with Headings, Paragraphs, Styling (Bold, Italic, Underlined), Footnotes, Image Captions, and a Multi-Language Translation Menu (This article provides a comprehensive overview of the v1.28 architecture of the web2txt en Skills design under gdn086, its operational rules, the Skills Documentation Page link, and the Google Drive folder information containing the SKILL.md file that anyone can download and directly use in their own system.)
Aydın Tiryaki
One of the most frequent obstacles encountered when reading, quoting, or archiving online articles, essays, academic papers, or news reports is how the core writing becomes surrounded by advertisement banners, navigation menus, sidebars, footers, cookie pop-ups, and social media widgets. When copying text from a webpage into a working document, structural problems such as broken paragraph layouts, unintended mid-sentence line wraps, and footnotes or photo captions mixed directly into body paragraphs routinely disrupt readability. Developed to make this extraction process clean, streamlined, and reliable, web2txt en (Skills Edition) v1.28 (gdn086) is a specialized content assistant that strips away all non-text surrounding clutter from shared URLs or researched topics and transfers the complete article with its original structural hierarchy intact.
The highest-priority rule governing the system is the “Never Abbreviate and Continuous Full Text Transfer” principle. Regardless of how long the source text is, no summarization, pruning, or artificial multi-part segmentation is applied; the entire text is transferred verbatim in a single uninterrupted flow. While all advertisements and surrounding page elements are completely ignored, the internal architecture of the article is faithfully preserved. A clear metadata header containing the original page title, author (if available), and publication date is placed at the very top, and all main headings and subheadings are preserved in their original hierarchy. Unnecessary line breaks caused by source code formatting are repaired so that every paragraph is presented as a continuous, unified block matching the original article.
A defining feature of the architecture is the preservation of inline typographic emphasis alongside the structured grouping of supplementary elements. Typographic attributes in the original writing—specifically bold, italic, and underlined text—are carefully preserved and reflected in the transferred output, while distracting inline redirection or ad hyperlinks are stripped away. Table data is transferred in a well-organized text layout preserving row and column hierarchy. Rather than cluttering the body text with inline image placeholder tags, the system separates all footnotes and presents them neatly and sequentially under a dedicated “Footnotes” heading immediately following the main text, followed by all photo and illustration descriptions gathered neatly under a dedicated “Image Captions” heading.
When a user specifies a topic instead of sharing a URL directly, web2txt en searches the internet and presents the best candidate webpages as an emoji-numbered list (1️⃣, 2️⃣, 3️⃣, etc.) featuring each page’s title, its directly clickable open URL, and a concise summary of its contents, asking the user which source they would like to extract. If a target page is blocked by a login wall, subscription paywall, or Cloudflare protection, the assistant clearly explains the technical restriction. Strictly when a webpage is completely closed or offline and may exist in the Wayback Machine (archive.org), the system scans archive.org snapshots, lists available records with emoji numbers, and transfers the selected archive version with a transparency notice and open archive URL. Multiple URLs are also processed cleanly without mixing sources.
With version v1.28, the system introduces an interactive multi-language translation menu immediately after the full text is extracted and displayed. Once the complete text—including headings, paragraphs, tables, “Footnotes”, “Image Captions”, open Source URL, and exact Python character count—is presented, the assistant asks: “🌐Would you like to translate this text into another language? | 1️⃣ 🇹🇷 TR Turkish | 2️⃣ 🇬🇧 EN English | 3️⃣ 🇩🇪 DE German | 4️⃣ 🇫🇷 FR French | 5️⃣ 🌍 Other | 🧭You can ask for help/guide at any stage.” When the user selects one of these options or specifies another language via “Other”, the entire extracted text is translated verbatim into the target language while preserving all structural headings, paragraphs, bold/italic/underlined styling, footnotes, and image captions, followed by a recalculated Python character count.
Maintaining full UTF-8 support, adhering to the official spelling Türkiye when referring to the Republic of Türkiye in English texts, and providing an open Source URL proof, exact Python (Interpreter) character counting, a Trust Score (1–10) for multi-source comparisons, double-layered data verification, a 3-level Living Assistant guide (1️⃣ 📝 Summary, 2️⃣ 🗂️ Index, 3️⃣ 📖 Detailed), and a 9-option visual/video generation menu presented at the very end after all operations conclude, the technical specifications of web2txt en are as follows:
Product Name: web2txt en (Skills Edition) Version and Lifecycle Status: v1.28 │ gdn086 │ GemStat:1 (In Development) Character Distribution: Title: 10 | Description: 984 | Instructions: 15121 | Total: 16115 characters 📄Skills Documentation Page: https://aydintiryaki.org/2026/06/19/gdn086/
Google Drive Folder Link and SKILL.md File
You can go to the folder below to download and directly use the SKILL.md file from this folder. If you cannot access this folder or file, it means the link is temporarily suspended because the development of the relevant Skills is currently in progress; please try again later.
📁 Google Drive Folder Path (EN): https://drive.google.com/drive/folders/1aUR3Cwnua6WPvInaK0kla3ZUiR79g_Kj
Note: You can make your own developments and edits on the SKILL.md file you obtain from here. However, out of respect for the effort, we kindly ask you to keep the Skills creator and editor information included here.
The design of web2txt en (Skills Edition) introduced in this article, as well as all content, perspectives, and views expressed herein, belong entirely to the author Aydın Tiryaki; system development was carried out with ATG Gem Factory (Skills Edition) v14.16, cloud file deployment with ATG7 Drive Workshop (Skills Edition) v01.07, and documentation packaging with ATG6 Document Workshop (Skills Edition) v01.20 standards, while editorial and writing assistance was performed by the active Gemini Pro model at the Extended thinking level (03.11.2026 14:18).
| aydintiryaki.org | YouTube | Aydın Tiryaki’nin Yazıları ve Videoları │Articles and Videos by Aydın Tiryaki | Bilgi Merkezi│Knowledge Hub | ░ Virgülüne Dokunmadan │ Verbatim ░ | ░ web2txt-tr │web2txt en (Gemini Skills) ░ 03.11.2026
