Privacy

Context Pack Privacy

Last updated: August 9, 2026

Context Pack handles the contents of PDF, DOCX, Markdown, and text files you choose. After you explicitly choose Capture page or Capture selection, it also handles readable website content, the page title, links, and a sanitized source URL. This data is used only to build the context pack you requested.

Supported documents, captured pages and selections, Markdown normalization, and token estimation are processed inside the Chrome extension.

What stays local

• Document and captured-page contents are not uploaded.
• There is no account, sync, analytics, advertising, or tracking integration.
• Working content, URLs, tab IDs, source lists, and generated Markdown remain in the open workspace and are not written to browser storage. Closing or reloading always discards the editable session. Copying or downloading keeps only the exported Markdown, not that session.
• After the first-run guide is closed, one non-content local preference prevents it from reopening automatically. It contains only the guide-completed value, can be cleared with browser data, and is not used for tracking or sync. Help can reopen the guide without changing the current pack.
• Clipboard content and downloaded Markdown created at your request remain under your control.
• Copy and download happen only after you choose them.

Page and selection access

The extension requests exactly activeTab and scripting. Chrome grants temporary activeTab access to the current page when you invoke the extension. Open current pack opens a new empty workspace when none is loaded or focuses the current workspace, without reading a page. Capture page and Capture selection explicitly open or focus that workspace and read only the invoking tab's main frame for local review. When no capture, review, or conversion is pending, later invocations from other tabs target the same current workspace; a busy attempt does not replace current work. Context Pack requests no host, browsing-history, storage, account, background, or persistent-content-script permission.

After you request page capture, a local safety preflight traverses main-frame DOM text and attributes and measures the UTF-16 length of current input and textarea values before cloning. Chrome copies that live form state during cloning, so this measurement is needed to reject oversized pages safely. Form values and interactive controls are omitted and never enter review, retained source state, or export. For a visible switch, checkbox, or radio control, only its bounded accessible label, non-value state such as on/off, checked, mixed, or selected, and whether it is actually disabled may be converted to plain text, including when the control appears inside an omitted form.

Limited Use

Information received through Chrome APIs is used only to provide the page or selection capture you request. It is not transferred, used for advertising, or made available for human review. This use follows the Chrome Web Store User Data Policy, including its Limited Use requirements.

Packaged processing

PDF.js, PDF workers and data assets, Mammoth, the DOCX worker, Markdown converters, browser capture, and the o200k_base tokenizer are packaged with the extension. Its policy limits connections and workers to the extension's own origin.

Website hosting

This page describes the extension. The website itself contains no analytics or advertising scripts; its hosting provider may process basic request metadata to deliver and protect the site.

Contact

Email support@tryaiblock.com.