Privacy Policy

OutlineMySource · Browser extension · Effective date: 6 May 2026

This policy describes how the OutlineMySource extension (“the Extension”) handles information when you use it. The Extension is built for a local-first workflow: outlining and transcription run on your device inside Chromium-based browsers, not on servers operated by the developer for processing your page text or media files.

Who we are

OutlineMySource is distributed as part of the AI4Context extension family. This policy covers the Extension as published in browser stores. For general questions about AI4Context, see ai4context.com.

What the Extension does

The Extension helps you generate a Markdown outline from:

• the active tab (article/main content where possible),
• local files you choose (e.g. PDF, plain text, subtitles), or
• pasted plain text in the panel.

For local audio or video, optional speech recognition and optional on-device semantic grouping may use machine-learning models loaded in the browser. See “Machine-learning assets” below.

Data we do not collect

The Extension does not require an account. We do not operate a backend that receives your outlined page content, pasted text, or local media files for processing on our servers as part of the core feature set described here.

Local storage in your browser

The Extension may use Chromium storage APIs (for example chrome.storage) to save small settings such as:

• UI language preference,
• identifiers needed to reuse the floating panel window instead of opening duplicates.

This data stays on your device unless you use browser sync features controlled entirely by your browser vendor.

Permissions (why they exist)

• Tabs / scripting / host access: used only when you start a capture on the current tab, so the packaged content script can read visible content from that page to build an outline. Broad host patterns reflect that users may visit many HTTP(S) sites; the Extension does not load arbitrary remote scripts into pages.
• Windows: used to open and focus the Extension’s own floating panel window.
• Storage: as described above.

Machine-learning assets (optional downloads)

When you use features that depend on models (for example local speech-to-text or embedding-based grouping), the Extension may download model weights and WebAssembly (ONNX) runtimes from public CDNs or model hubs (such as Hugging Face / jsDelivr) into your browser cache. Those downloads are standard for in-browser ML and are not used to upload your content to the developer.

After download, inference runs locally in your browser subject to the Extension’s content security policy.

Exports and downloads

When you export Markdown, the file is created through your browser’s normal download flow and saved where your browser or operating system directs downloads. The Extension does not send that file to our servers.

Changes

We may update this policy when the Extension changes materially. The “Effective date” at the top will be revised when this page is updated.

Contact

For privacy questions about this Extension, open an issue on the project repository or use the contact options listed on ai4context.com.

This page is intended for store listings (Chrome Web Store, Microsoft Edge Add-ons, etc.). If you enable GitHub Pages from the docs/ folder, ensure you only publish content you intend to be public.