yiidtw.dev / amem-clipper

amem Clipper — Privacy Policy

Last updated: 2026-08-22

amem is a local-first knowledge capture system. This document covers the Chrome extension ("amem Clipper") and the native CLI ("amem").

What amem does

amem captures web pages, papers, and recordings into your personal, local-first knowledge base. Everything stays on your machine unless you explicitly opt into a sync destination.

Data collection

amem does not collect, transmit, or store any user data on external servers. There is no backend, no database, no analytics, no telemetry.

What amem Clipper accesses

Captures start in one of two ways, and both are worth stating plainly:

You ask. Clicking the amem action, opening the side panel, or picking a context-menu entry reads the page title, URL, and visible text.
An AI agent on this machine asks. amem Clipper's purpose is to act as the browser for a local agent — for example Claude Code running in your terminal. That agent sends commands (navigate, click, read this page) to the extension over a bridge on 127.0.0.1, and pages read that way can be written to your knowledge base without a click in the browser. The agent runs on your machine, under your account, at your instruction; no remote party can reach this bridge.

Also accessed:

Local storage (chrome.storage.local) — for capture history and preferences.
Local HTTP bridge at http://127.0.0.1:7601 — the only channel between the extension and the amem daemon on your machine, used to write captures to your filesystem. Bound to loopback; no external network calls.
Tab recording (tabCapture) — only while you are actively recording from the side panel. Recordings are saved directly to your browser Downloads folder as .webm files.

What gets written, and how to change it

The extension does not decide this. It reports what it read to the amem daemon, and the daemon decides what to keep, using ~/.amem/digest.json on your own disk:

keep_verbs — which kinds of read are treated as content at all. Anything not listed is never written.
min_chars — a length floor, so button labels and prices do not become notes.
deny_hosts — host substrings you want skipped entirely. Empty by default. Earlier versions shipped a guessed list of banking and webmail hosts; it was removed in favour of a setting you control, because a list we guess at cannot cover your banks and quietly mis-fires on ordinary sites.

Edit that file and the change takes effect on the next capture — no extension update, no reinstall.

What amem Clipper does NOT do

Does not read your browsing history.
Does not track your activity or behavior.
Does not send data to any third-party server.
Does not use Google Drive, OAuth, or any cloud provider in this release.
Does not record audio.

Permissions explained

Permission	Why
`storage`	Save capture history and settings locally in the browser.
`activeTab`	Access the current tab when you initiate a capture.
`scripting`	Inject capture helpers into the active tab when you press capture.
`sidePanel`	Render the amem side panel UI.
`alarms`	Periodically retry the local bridge connection.
`tabs`	Read the current tab's URL and title for the capture record.
`tabGroups`	Name and colour tab groups when you use the sort/cluster-tabs feature.
`tabCapture`	Record the active tab's video stream (explicit user action only).
`offscreen`	Host `MediaRecorder` in an offscreen document (service workers can't).
`contextMenus`	Offer right-click capture entries.
`downloads`	Save recordings to your Downloads folder as `.webm` files.
`favicon`	Show each captured page's site icon in the side panel list.
`host_permissions: <all_urls>`	Capture works on whatever page you are viewing, and lets a local agent read a page you point it at. Broad by necessity: the agent decides which page to open, so the set cannot be enumerated in advance. Nothing is read except in response to an explicit request — yours, or your local agent's.

amem CLI + crate

The native amem binary (installed via amem install, brew, or cargo) runs entirely on your machine. It reads and writes only ~/.amem/ and (optionally) paths you pass as arguments. It makes outbound HTTPS requests only when you run amem capture <url> — strictly to fetch the URL you asked for (e.g., arXiv PDF, YouTube audio via yt-dlp). No analytics, no telemetry.

Contact

Questions about this policy, or any request concerning your data: email yiidtw@gmail.com. A privacy policy has to name a channel that actually reaches someone, so this one is a mailbox rather than an issue tracker.