# Ambient Context: Automatically Track Your Work History

This MacOS app records your focused windows in markdown format for productivity tracking.

- **Source:** [Hacker News: Front Page](https://brifing.si/source/b07aa772-b1a0-4a76-aa18-719c6f9e2dde)
- **Published:** 2026-08-25
- **Topic:** [Product Management](https://brifing.si/c/product)
- **Original article:** <https://github.com/dragthelake/ambient-context>
- **On Brifing:** https://brifing.si/s/e96eb650-d809-4f7b-9aa2-78ceb7d749df

## What is Ambient Context?

Ambient Context is a MacOS menu bar application that monitors your focused windows. It reads the text content every few seconds and saves it in markdown format. The app creates daily files that help users track their work history. Currently in early development, it requires manual installation by users. The tool is particularly useful for developers and knowledge workers.

## How It Works

The application uses MacOS's accessibility tree to read text from the currently focused window. It appends this data to daily markdown files in a user-selected folder. These files can later be fed to AI assistants like Claude Code to answer questions about past work. Recording only happens with active windows when the screen is unlocked, ensuring privacy. The app skips password managers and private browsing sessions automatically.

## Installation Process

As an unsigned application currently, users need to build it themselves. Installation requires Node.js, Rust and Xcode Command Line Tools. The project can be cloned from GitHub and built in about 2 minutes. After installation, users must grant accessibility permissions and select a recording folder. The app starts automatically after initial setup and maintains recording state between sessions.

## Privacy and Security

Ambient Context only reads the currently focused window, ignoring background or minimized windows. It automatically skips password managers and private browsing tabs. Secure input fields are blocked at the accessibility level. All data remains as plain text on the user's local disk with no cloud synchronization. The recording folder is excluded from self-observation to prevent recursive recording.

## Compatibility Notes

Chromium and Electron-based apps (like Chrome, VS Code, Slack) build their accessibility trees only when requested, causing initial captures to be incomplete. GPU-rendered terminals (Kitty, Alacritty) expose little text while Terminal.app and iTerm2 work fully. Chrome may show minor window resize animation glitches during recording - a known side effect of the capture mechanism.

## Use Cases

Users can automatically document their daily work routines, build project memory, or generate standup reports. Particularly valuable for developers, the tool makes it easy to track time spent on different projects. The recorded files can be processed by LLMs to generate personalized reports. Right-clicking the menu bar icon provides quick access to today's file, the folder, and settings.

## Technical Details

Built using the Tauri framework with Rust backend, the app creates daily markdown files with timestamped headings showing the day's timeline. Default storage is set outside ~/Documents to avoid iCloud syncing. The eye icon in the menu bar indicates recording status - open eye means active, closed means paused. Body lines are written once per day regardless of how often they're seen to keep files compact for LLM processing.

## Future Potential

Ambient Context fills an important gap in digital productivity tools by automatically documenting work context. It helps developers analyze work patterns and improve time management. Wider adoption is expected after Apple Developer enrollment completes. Future updates could expand support to other operating systems and applications based on user feedback, making it a versatile work memory assistant.

## Sources covering this story

- https://github.com/dragthelake/ambient-context — 

---

_Summary written by Brifing from the original reporting. Copyright in the original article belongs to its publisher; read the full piece at the link above._
