Frequently Asked Questions (FAQ)

June 1, 2026 · View on GitHub

Common questions about iDO and how to use it.

iDO is a local-first AI desktop copilot that captures your computer activities, uses LLMs to understand context, and recommends tasks based on what you're doing. Think of it as an AI assistant that watches what you do and helps you stay organized.

Is iDO free?

iDO itself is free and open source. However, you need to provide your own LLM API key (e.g., OpenAI), which has usage costs. Typical costs are $0.01-0.10 per hour of activity, depending on the model you choose.

Which platforms are supported?

Currently:

✅ macOS 13 (Ventura) or later
✅ Windows 10 or later
✅ Linux (Ubuntu 20.04+ or equivalent)

Is my data private?

Yes! iDO is designed with privacy as a priority:

✅ All data stored locally on your device
✅ No cloud uploads or syncing
✅ You control your own LLM API key
✅ Open source code (audit it yourself)

The only data that leaves your device is what's sent to your LLM provider (OpenAI or compatible) for analysis.

Installation & Setup

Where can I download iDO?

Download the latest version from GitHub Releases: 👉 https://github.com/UbiquantAI/iDO/releases/latest

How do I install on macOS?

Download the .dmg file for your Mac (Intel or Apple Silicon)
Open the .dmg and drag iDO to Applications
Launch iDO from Applications
Grant permissions when prompted

See the Installation Guide for details.

Why does macOS block iDO from running?

macOS blocks apps downloaded from the internet as a security measure. To allow iDO:

Go to System Settings → Privacy & Security
Find "iDO was blocked from use"
Click Open Anyway
Confirm

What permissions does iDO need?

On macOS, iDO requires:

Accessibility - To monitor keyboard and mouse events Screen Recording - To capture screenshots

You'll be prompted to grant these on first run. See Installation Guide for details.

Do I need an OpenAI API key?

Yes, iDO requires an LLM provider to analyze activities. The recommended provider is OpenAI, but any OpenAI-compatible API works (e.g., Azure OpenAI, local LLM servers).

Get an OpenAI API key at: https://platform.openai.com/api-keys

Which LLM model should I use?

Recommended:

gpt-4o-mini - Best value, good quality, ~$0.01 per hour (default)
gpt-4o - Higher quality, ~$0.05-0.10 per hour

Tips:

Start with gpt-4o-mini (default) for best value
Upgrade to gpt-4o if you need higher quality
You can change models anytime in Settings

Usage

How does iDO track my activities?

iDO uses a three-layer approach:

Perception Layer: Captures keyboard, mouse, and screenshots (default: 0.2 seconds)
Processing Layer: Filters noise and uses LLM to create meaningful activity summaries
Consumption Layer: Displays activities and generates task recommendations

Will iDO slow down my computer?

iDO is designed to be lightweight:

CPU: ~2-5% usage during capture
RAM: ~200-500 MB
Disk: ~100-500 MB per day

If you notice high resource usage, try:

Increasing capture interval (Settings → 2-3 seconds)
Lowering image quality
Disabling unused monitors

How much disk space does iDO use?

Storage varies based on your settings:

Typical usage (default settings):

~100-300 MB per day
~3-9 GB per month

Factors:

Capture interval (0.2s = 5 screenshots/sec/monitor)
Image quality (85% is default)
Number of monitors
Image optimization (reduces duplicates)

Can I exclude certain screens or apps?

Monitors: Yes! Go to Settings → Screen Capture and disable specific monitors.

Apps: Not yet, but this feature is planned.

Workaround: Stop capture (Dashboard → Stop) when using sensitive apps, then resume.

What happens to old data?

Currently: Data is kept indefinitely until you manually delete it.

Coming soon: Automatic cleanup options:

Delete screenshots older than X days
Delete completed tasks older than X days
Keep database but remove old screenshots

Can I export my data?

Coming soon: Export features for activities and screenshots.

Currently: You can access the raw database at:

macOS: ~/.config/ido/ido.db (SQLite format)

Troubleshooting

iDO isn't capturing any activities

Check:

✅ Permissions granted (Accessibility + Screen Recording)
✅ At least one monitor enabled in Settings
✅ Check system status in sidebar

Solutions:

Restart iDO
Re-grant permissions (System Settings → Privacy & Security)
Check logs: ~/.config/ido/logs/app.log

LLM connection test fails

Common causes:

❌ Invalid API key
❌ No internet connection
❌ API endpoint unreachable

Solutions:

Verify API key is correct (check for extra spaces)
Test internet connection
Try a different model
Check OpenAI service status: https://status.openai.com/

Screenshots aren't being captured

Check:

✅ Screen Recording permission granted
✅ Monitor is enabled in Settings → Screen Capture
✅ Disk has free space

Solutions:

Re-grant Screen Recording permission
Restart iDO
Check screenshot folder: ~/.config/ido/screenshots/

Activities have generic titles like "Activity 1"

Cause: LLM isn't generating proper summaries

Solutions:

Verify LLM connection (Settings → Test Connection)
Check API key has available credits
Try a different model (gpt-4o vs gpt-4o-mini)
Review logs for errors

iDO is using too much CPU/RAM

Solutions:

Increase capture interval:
- Settings → Screen Capture → Capture Interval: 2-3 seconds
Lower image quality:
- Settings → Screen Capture → Image Quality: 70%
Disable unused monitors:
- Settings → Screen Capture → Uncheck monitors
Enable image optimization:
- Settings → Screen Capture → Image Optimization: On

App crashes on startup

Solutions:

Check logs: ~/.config/ido/logs/app.log

Try resetting settings:

mv ~/.config/ido/config.toml ~/.config/ido/config.toml.backup

Reinstall iDO
Report issue with logs: https://github.com/UbiquantAI/iDO/issues

Privacy & Security

What data does iDO collect?

iDO captures:

Keyboard and mouse events (timestamps, event types)
Screenshots (periodic captures)
Window titles and application names
Activity timestamps and durations

Not captured:

Raw keystroke content (only event summaries)
Passwords or secure fields
Audio or video

What gets sent to OpenAI?

When processing activities, iDO sends:

Screenshots (as base64 data)
Event summaries (e.g., "typing in VSCode")
Timestamps and window titles

Not sent:

Your raw database
Complete keystroke logs
Data from disabled monitors

Can my company use iDO?

Things to consider:

iDO captures screenshots of everything on your screen
Data is sent to your LLM provider (OpenAI, etc.)
Data is stored unencrypted locally

Recommendations:

Check your company's security policy
Use a company-approved LLM provider
Consider using Azure OpenAI for enterprise deployments
Store the database on an encrypted volume
Don't capture screens with sensitive information

Is iDO open source?

Yes! iDO is licensed under Apache 2.0.

View the source code: https://github.com/UbiquantAI/iDO

How is iDO different from Rewind.ai?

Similarities:

Both capture screen activity
Both use AI to analyze activities

Differences:

Feature	iDO	Rewind.ai
Source	Open source	Closed source
Privacy	Local-only	Cloud option
LLM	Bring your own	Built-in
Cost	Free (+ API costs)	Subscription
Platform	macOS, Windows, Linux	macOS only