You worked hard today. But do you actually know where the time went?
Most of us don't. We remember the feeling of a busy day — the tabs, the switching, the "just one more thing" — but when someone asks what we shipped, or why the week felt wasted, the honest answer is a guess.
DayTrail is the app that tells you the truth about your own time. Quietly, privately, entirely on your machine — it builds a real record of your day from the apps you use, the projects you work in, the tabs you visit, the AI tools you lean on. No timers to start. No notes to write. No data leaving your computer.
At the end of the day, you'll know exactly where the hours went. And slowly, that knowledge changes everything.
macOS Apple Silicon — Homebrew recommended:
brew tap varaprasadreddy9676/tap && brew trust varaprasadreddy9676/tap && brew install --cask daytrailmacOS — one-line installer, no Homebrew needed:
curl -fsSL https://raw.githubusercontent.com/varaprasadreddy9676/DayTrail/main/scripts/install-macos.sh | bashWindows: download the latest .msi or .exe from the Releases page.
If macOS or Windows warns because the app is not code-signed yet, see Troubleshooting. DayTrail is open source, local-first, and built from this repository.
It's 6pm. You're tired. You're pretty sure you worked hard — you had your coffee, opened your laptop, kept things moving. But you can't really account for the day. There was a PR review, some Slack, a bug you got pulled into, a YouTube video you "just needed a minute" for. Your standup tomorrow is going to be vague. Your timesheet is going to be a reconstruction. You'll say you spent time on the project when really you spent time near it.
This isn't a discipline problem. Human memory is simply not built for tracking time. We remember events, not durations. We remember the last thing, not the whole arc. We remember effort as output — and when the output was invisible (reading, researching, getting unstuck), we forget it happened at all.
DayTrail closes that gap. Not by watching you, but by remembering what your computer already knows.
| What matters | DayTrail | Traditional time trackers | Cloud AI trackers |
|---|---|---|---|
| Privacy | Local-first, no account, no backend | Often local, but manual | Usually cloud-based |
| Effort | No timers, no manual notes | Start/stop timers and edits | Passive, but opaque |
| Footprint | Small Tauri + Rust desktop app | Often heavier desktop stacks | Higher CPU/RAM from rich capture |
| AI | Optional, BYO key or local Ollama | Usually none | Built in, usually subscription-based |
| Developer context | Apps, tabs, projects, files, terminals, AI tools | Mostly app names and categories | General activity, often screenshot-heavy |
DayTrail is not trying to be surveillance software. It is trying to be a private work memory: enough context to reconstruct your day, without sending your life to someone else's server.
One view. No setup required. You'll see what you're doing right now (not just "Chrome" — the actual tab, the actual project), the real breakdown of where your hours went, and a timeline you can scroll through hour by hour.
The first time you see it, you'll feel something. Maybe recognition. Maybe a little discomfort. Both are useful.
Where your day actually went Not a rough estimate — the real thing. The 90 minutes in email you'd have called "a quick check." The hour on YouTube you'd have written off as five minutes. When you see it clearly, you can change it.
Whether you can trust your own memory Compare what you remember with the captured timeline. Most people are surprised. Not because they're lazy — because memory is unreliable about time. Once you know this about yourself, you stop blaming yourself for "feeling" productive while getting less done than expected.
Where focus disappears DayTrail shows you the exact moments your attention fragmented — context switches, tab jumps, the gap between intention and action. Patterns repeat. Once you see yours, they're hard to unsee.
How much of your work runs through AI ChatGPT, Claude, Copilot, Cursor — DayTrail tracks all of them as first-class work. You'll know which tools you actually rely on, for how long, and on which projects. As AI becomes woven into how we work, this matters more every week.
What to say at your standup Source-backed. Specific. No reconstructing from memory the night before.
Built with Tauri + Rust — no bundled Chromium runtime. DayTrail runs quietly in the background, keeps installer sizes small, and stays out of your way while it captures the context you would otherwise forget.
Start a focus block — 25, 50, or 90 minutes — and DayTrail sends a gentle native notification when you've drifted to YouTube, Reddit, WhatsApp, or other distractions. It reminds you. It never blocks apps. It never judges. And every session is saved, so you can review later whether the block actually stayed on track.
The difference between a nudge at minute three and discovering the drift at hour two is enormous.
DayTrail does not just log data. With Claude, GPT-4, Gemini, DeepSeek, or a local Ollama model configured, it helps you reason about what happened:
- Ask AI: ask plain-English questions like "Which projects did I ignore this week?" or "What did I spend the most time on today?"
- Proactive insights: background analysis surfaces patterns you would not have thought to search for.
- AI impact tracking: ChatGPT, Claude, Codex, Copilot, Cursor, and similar tools become first-class work sessions instead of disappearing into "browser time."
- Daily and weekly reports: turn captured work into a standup, timesheet, client update, or weekly retro draft.
Because the provider is optional and configured locally, you choose whether DayTrail stays fully offline or uses your own AI key.
Examples of insights DayTrail can surface:
- You've had three days without a real deep work block
- Your context-switching spiked 60% this week compared to last
- You have two open commitments that haven't been touched in four days
- Your AI tool usage doubled but the project output didn't
High-priority insights fire an OS notification. All of them live in the Insights view — dismissable, filterable, with a one-click "Explore in chat" button that takes you directly into a conversation about what was found.
Native notifications mean the useful patterns come to you while the day is still happening — not only after you remember to open a dashboard.
This is what makes DayTrail feel like an AI-native app rather than a tracker with a dashboard.
DayTrail also tracks which AI tools you actually rely on, for how long, and on which projects — because AI work is real work, not random browser time.
Enabled optionally in Settings. When turned on, DayTrail watches the same foreground-window signals it already uses and notices when you've been at it for a while. It sends blink reminders, posture resets, and short break prompts — at the interval you choose. It stays quiet during calls, presentation-like contexts, or when you step away. No extra card on your Today screen. No medical claims. Just the kind of nudge a good colleague might give you.
After an interruption — a meeting, a lunch, an unexpected call — DayTrail shows you what you were in before you left. The app, the project, the file, the context. You don't have to rebuild the mental model from scratch. The trail is there.
One click generates a source-backed report of everything you worked on — sessions, apps, projects, AI tools used, and items to review. With an AI provider connected, it turns the raw log into a first draft for your standup, client update, or weekly retro. You edit; you don't invent.
Every work session breaks down into the apps, projects, and AI tools behind it. A single block of time tells the whole story — not just "I worked on Project X" but how you moved through it.
Most trackers fail silently. You lose a full day of data before noticing. DayTrail watches its own capture engine — if a permission gets revoked or a bridge stops working, it tells you exactly what's wrong and how to fix it.
This is non-negotiable for us. Everything DayTrail captures stays on your machine. Always.
- No cloud sync. No account. No backend you're trusting someone else to secure.
- Screenshots are off by default.
- Clipboard content is never stored.
- Browser URLs are redacted before storage where possible.
- AI providers are optional, configured locally, and queried only when you ask.
- If you uninstall DayTrail, your data stays on your machine — in your control.
See PRIVACY.md for the complete model.
DayTrail also ships with DayTrail Helper, a read-only Codex plugin for people who want to ask questions about their captured work from inside Codex.
It connects to the local DayTrail SQLite database on your machine and can answer things like:
- "Summarize my DayTrail activity today."
- "Show my open DayTrail tasks."
- "Search DayTrail for Slack yesterday."
- "What AI tools did I use this week?"
- "Show my recent DayTrail reports."
The plugin does not upload, sync, or modify your data. It reads your local database in read-only mode through an MCP server.
See plugins/daytrail-helper for install and usage details.
Built with Tauri + Rust — no bundled Chromium runtime. DayTrail ships as a small native desktop app instead of bundling a whole browser engine. Current release installers are roughly 10-12 MB on macOS and 6-9 MB on Windows, depending on installer format.
The fastest install paths are also listed in Quick Start.
macOS Apple Silicon — Homebrew recommended:
brew tap varaprasadreddy9676/tap
brew trust varaprasadreddy9676/tap
brew install --cask daytrail
brew trustis required once. Recent versions of Homebrew require you to explicitly trust third-party taps before installing casks from them. You only need to run this once per machine.
To update to the latest version:
brew update && brew upgrade --cask daytrail
brew updatefirst is required.brew upgradealone uses your local tap cache — withoutbrew update, Homebrew won't know a new version exists and will say "already installed". Always run both commands together.
macOS — one-line installer (no Homebrew needed):
Paste this in Terminal — it downloads the latest release, installs to /Applications, and clears the Gatekeeper flag automatically:
curl -fsSL https://raw.githubusercontent.com/varaprasadreddy9676/DayTrail/main/scripts/install-macos.sh | bashmacOS — manual DMG: grab the latest .dmg from the Releases page, drag to Applications, then run this once to clear the Gatekeeper flag:
xattr -dr com.apple.quarantine /Applications/DayTrail.appWindows: download the .msi or .exe installer from the same Releases page.
Other Macs / build from source: see Try it below.
macOS: "DayTrail.app is damaged and can't be opened."
It is not damaged. Because the app isn't notarized (no paid Apple Developer ID), macOS blocks the first launch. There are three ways to fix this — pick the easiest:
Option 1 — use Homebrew (handles it automatically):
brew tap varaprasadreddy9676/tap
brew trust varaprasadreddy9676/tap
brew install --cask daytrailOption 2 — use the one-line installer (handles it automatically):
curl -fsSL https://raw.githubusercontent.com/varaprasadreddy9676/DayTrail/main/scripts/install-macos.sh | bashOption 3 — manual fix after DMG install (one command, one time per version):
xattr -dr com.apple.quarantine /Applications/DayTrail.appDrag the app to Applications first, then run the command above, then open normally.
Windows: "Windows protected your PC" (SmartScreen)
The installer isn't code-signed yet, so SmartScreen warns on first run. Click More info → Run anyway.
Homebrew: "Refusing to load cask from untrusted tap"
Recent versions of Homebrew require you to trust third-party taps before installing casks. Run once:
brew trust varaprasadreddy9676/tap
brew install --cask daytrailHomebrew: "already installed" or installs an old version
Homebrew caches tap metadata locally. If brew upgrade --cask daytrail says the latest is already installed but you know a newer version exists, your local cache is stale.
Fix:
brew update && brew upgrade --cask daytrailIf it still shows the wrong version, force a clean reinstall:
brew uninstall --cask daytrail
brew update
brew install --cask daytrailmacOS: capture stopped / titles show only the app name
DayTrail needs Accessibility permission to read window titles. Open Settings → Capture Health in the app — if Accessibility shows as missing, use Fix accessibility to re-grant it. A macOS update can silently reset this.
- Install the app and grant macOS Accessibility (or Windows equivalent) when prompted.
- Enable browser extension support if you want tab titles and domains.
- Install editor and terminal integrations for project and file context.
- Allow OS notifications for Focus Mode nudges, Smart Breaks, and proactive AI insights.
- Set your working hours in Settings → Capture Health so DayTrail never asks "were you away?" at midnight.
- Add an AI provider in Settings if you want generated digests, proactive insights, and Ask AI answers. Claude, GPT-4, Gemini, or a local Ollama model all work.
- Leave DayTrail running from startup. One full workday of capture is when it starts to get interesting.
| Platform | Status |
|---|---|
| macOS | Primary target. Fully exercised — install, permissions, tray, capture, reporting, Focus Mode, and AI flows. |
| Windows | Backend, tray, terminal bridge, credential storage, and CI installers are implemented. Real smoke testing still required before a signed public release. |
| Linux | Not a release target yet. Some Tauri pieces may work; capture behavior isn't validated. |
apps/desktop/ Tauri desktop app — Rust backend, React UI
apps/browser-extension/ Browser extension for tab context
apps/vscode-extension/ VS Code / editor bridge
plugins/daytrail-helper/ Codex plugin for read-only local DayTrail queries
.agents/plugins/ Repo-local Codex marketplace entry
scripts/ Build, release, bridge, and verification scripts
docs/ Supporting docs and screenshot assets
Requirements: Node.js 20+, npm, Rust stable, Tauri platform prerequisites for your OS.
npm ci --prefix apps/desktop
cd apps/desktop && npm run tauri devRun the full quality gate:
npm run release:checkTargeted checks:
npm run desktop:check
npm run desktop:test
npm run browser-extension:check
npm run vscode-extension:check
npm run test:scriptsmacOS unsigned local build:
npm run desktop:dmgWindows installer from a Windows machine:
npm run desktop:windowsSee RELEASE.md for the full release checklist, signing notes, and verification steps.
Every non-release push to main triggers a release candidate. If the commit bumps the desktop version, GitHub Actions tags that version. If the current version is already tagged, Actions bumps the patch version, commits, tags, and dispatches macOS and Windows builds automatically.
Use scripts/release.sh <version> to cut a specific version manually.
- PRIVACY.md — local storage, metadata capture, redaction, exports, AI provider behavior
- SECURITY.md — how to report security issues
- RELEASE.md — release verification checklist
- docs/screenshots/README.md — public screenshot set
DayTrail is free and open source, built by one developer in the open. If it's helped you understand your own days — or you just want to see it keep growing — you can support development here:
Stars, issues, and pull requests are equally welcome.
MIT OR Apache-2.0. See LICENSE, LICENSE-MIT, and LICENSE-APACHE.