Feature Guide

Owlfy Feature Guide

Product Overview

Owlfy is a next-generation Voice AI Operating System that redefines how humans interact with computers through full voice interaction. Command your computer with natural language and let AI seamlessly integrate into daily operations for 10x efficiency gains.

Product Slogan: Speak. It’s Done.

Quick Start

Voice Key Operations

Owlfy’s core interactions are completed through the voice key:

Operation	Function	Use Case
Click voice key	Launch “Voice Input”	Voice-to-text input
Hold voice key	Launch task execution	Sonic Execution (quick tasks) or AI Assistant (complex tasks)

Default Voice Keys:

Windows: Right Alt key or Mouse wheel
macOS: Fn key or Mouse wheel

You can customize the voice key in Settings.

Text Command Bar

When it’s inconvenient to speak, use the text command bar:

macOS: Option + Space
Windows: Ctrl + Space

Type your request and press Enter — equivalent to holding the voice key.

Core Features

1. Sonic Execution

Triggered by holding the voice key and speaking a quick command. Instantly completes computer operations.

1.1 Quick Launch

Say the name to instantly open apps, websites, or directories.

Type	Example	Result
Application	”Open WeChat”	Launches WeChat
Website	”Open GitHub”	Opens github.com
Directory	”Open Downloads”	Opens Downloads folder

Custom Aliases: Set your own aliases for apps, websites, and directories in Settings.

1.2 Instant Search

Say search content and specify a search engine for quick results.

Command	Effect
”Search for iPhone 16 review”	Opens default search engine
”Google search climate change”	Searches Google
”Bing search AI trends”	Searches Bing

1.3 Smart Text Processing

Select text, hold the voice key, and speak your need. AI processes instantly.

Text Polishing & Rewriting

Input	Command	Output
”Boss, I can’t do this project, it’s too hard."	"Rewrite this in a professional tone"	"I’m encountering some technical roadblocks with this project and would appreciate discussing potential resources or alternative approaches."
"Do it yourself."	"Make this sound more tactful"	"Could you please handle this yourself?"
"Had amazing hotpot today and it totally made my day."	"Turn this into an Instagram caption"	"Hotpot kind of night and absolutely no regrets. So good, so comforting. 🍲✨ #FoodieFinds #HotpotLove”

Information Extraction & Summarization

Input	Command	Output
”Let’s review the design at 10 AM tomorrow. Bob, print 3 copies. Alice, book room 2."	"List the action items”	• 10:00 AM: Design review • Bob: Print 3 copies • Alice: Book Room 2
”Q3 revenue increased 20% year over year thanks to strong cloud growth, but margins dipped slightly because of higher R&D spending."	"Make this shorter"	"Cloud growth lifted Q3 revenue, but higher R&D spending hurt margins.”

Code Processing

Input	Command	Output
`const sum = arr.reduce((a, b) => a + b, 0);`	”Explain this code"	"This uses the reduce method to iterate over ‘arr’, accumulating the sum of its elements starting from 0.”
`const sum = arr.reduce((a, b) => a + b, 0);`	”Turn this JavaScript into Python”	`sum_result = sum(arr)`

Knowledge Queries

Input	Command	Output
”FOMO"	"What does this mean?"	"Fear of missing out — the anxious feeling that other people may be having rewarding experiences without you."
"Despite of the heavy rain, but we still arrived on time for the meeting."	"Fix the grammar"	"Despite the heavy rain, we still arrived on time for the meeting.”

Data Transformation

Input	Command	Output
”Bob expensed $150 for taxi, Alice spent $200 on meals, and John paid $80 for taxi."	"Turn this into a table”	\| Name \| Type \| Amount \| \| --- \| --- \| --- \| \| Bob \| Taxi \| $150 \| \| Alice \| Meals \| $200 \| \| John \| Taxi \| $80 \|

1.4 Smart File Processing

Select one or more files, hold the voice key, and speak your requirements. AI quickly processes them.

Image Processing: Compress, convert formats, remove backgrounds, adjust resolution, OCR, flip, rotate, crop, color adjustment, multiple images to GIF/PDF.

Audio/Video Processing: Compress, add watermark, speed up/slow down, smart silence removal, concatenate, format conversion, extract audio, frame extraction, voice recognition.

Document Processing: Merge/split PDF, PDF watermark, PDF to image, Markdown to PDF/Word, draw charts, edit Excel/Word/PPT, AI generate PPT, document to Markdown.

2. Voice Input

Triggered by clicking the voice key. Not just transcription — an intelligent text input experience.

Input Flow

Click the voice key → Start speaking
Click again to end
Owlfy applies role polishing
Output appears at the current cursor position

Role Polishing

Choose from multiple roles to polish your voice input:

Role	Effect
Oral Correction	Converts spoken language into formal written language
Workplace Professional	Rewrites casual speech into polished, high-EQ workplace expressions
Custom Role	Create your own roles for personalized needs

Example:

Say: “Free for dinner tomorrow?”
Select “Workplace Professional” → Output: “Would you be available for dinner tomorrow evening?”

Multi-language Output

Specify the output language to achieve real-time translation while recording.

Example:

Say in Chinese: “明天一起开会讨论一下这个方案”
Set output language to English → Output: “Let’s meet tomorrow to discuss this proposal.”

Smart Snippets

Map long text to short phrases for quick input.

Example:

Set: “123 Main Street, New York, NY 10001, (555) 123-4567” = “Shipping address”
Say: “Shipping address” → Full address and contact info is entered instantly

Privacy Guarantee: Snippets are stored locally with no network transmission.

3. AI Assistant

Triggered by holding the voice key and describing a complex task. Owlfy has a powerful AI scheduling engine that completes tasks through planned execution.

Full Capability Support

Intelligently invokes a vast array of Skills to handle:

Images, Audio & Video: Format conversion, compression, cropping, smart editing, material extraction, dubbing, etc.
Documents: Writing, summarizing, translating, rewriting, merging, splitting, format conversion, etc.
System Operations: Organizing files, cleaning junk, shutdown/lock screen, etc.
Web Automation: Searching, extracting web content, downloading resources, automating web operations, etc.

Plan & Execute

Breaks large tasks into executable plans, then tackles them one by one through code, CLI commands, APIs, MCP, or Skills — delivering the final result.

Example:

Task: “Summarize the key points from all PDFs in my Downloads folder and create a Word document”
AI Assistant plan:
1. Scan Downloads folder for PDF files
2. Extract text from each PDF
3. Summarize key points
4. Generate a formatted Word document

Local-First

For tasks that can be completed locally through programs, scripts, or tools — such as document processing, image processing (non-AIGC), and audio/video processing (non-AIGC) — AI Assistant prioritizes local execution. Files are not uploaded to the cloud, ensuring data security.

4. Scheduled Tasks

Let your computer learn to work “automatically and spontaneously.”

Periodic Execution

Supports one-time or recurring tasks.

Example Scenarios:

Check and summarize important emails every hour
Auto shutdown at 21:00 every day
Every Monday at 9:00 AM: Summarize last week’s meeting notes and create this week’s to-do list

5. Remote Control

Break spatial limits. Send tasks via mobile instant messaging apps (such as WhatsApp, Messenger). The Owlfy assistant completes tasks on your controlled computer and feeds results back to your phone in real time.

Settings

General

Setting	Description
Run on system startup	Start Owlfy automatically when the computer boots
Mouse wheel as voice key	Use mouse middle button as the voice shortcut
Keyboard voice key	Set which key triggers voice functions (Mac default: Fn, Windows default: Right Alt)
Text command bar shortcut	Set shortcut for the text command bar
Screenshot shortcut	Set shortcut for the screenshot tool
Interface language	Switch software display language
Appearance theme	Light, Dark, or follow system
System permissions	View and enable Accessibility, Screen Recording, Full Disk Access (mainly for macOS)

Clipboard

Setting	Description
Clipboard shortcut	Set shortcut to open clipboard history
History count	View current clipboard history capacity

AI Assistant

Setting	Description
AI Assistant sandbox mode	Determines the scope of operations AI can perform on your computer
AI Assistant Workspace	In restricted mode, AI can only read/write files in the specified folder
AI Assistant approval policy	Whether AI asks for confirmation before high-risk or restricted operations
Confirm before AI Assistant tasks	When enabled, tasks sent to AI Assistant require confirmation before execution

Voice Input

Setting	Description
Microphone	Select recording device and test microphone
Mute system during recording	Automatically mute system sound during voice input to avoid interference
Voice bar sound effects	Play sounds when voice bar appears and disappears

Account

Setting	Description
Login/Logout	Manage account login status
Membership info	View membership status, validity, and renewal options
My devices	View devices logged into your account
Edit device name	Rename devices for easier identification
Refresh device list	Re-fetch latest device status
Remove device	Remove a specific device from your account
Redemption code	Enter codes to redeem points or membership benefits

About

Setting	Description
Current version	View version number
Check for updates	Manually check for new versions
Install update	Install when a new version is downloaded
Official homepage / social media	Links to official website and social accounts

Privacy & Security

Owlfy follows the principle of “Local-First, On-Demand, No Excessive Retention.”

1. Local Personalized Configuration

Your personal settings are saved locally, not stored long-term on servers.

Local content: Aliases, smart snippets, custom roles, etc.
More controllable: Content follows the current device, not becoming a platform-side long-term profile.
Device change reminder: Re-setup is usually needed when changing computers.

2. Local History

Usage history is kept locally by default.

Local-first: Chat history and operation logs are saved only on your device.
Easy to trace: Conveniently review past actions and conversations.
Reduced leakage: History does not automatically become long-term cloud archives.

3. Voice Privacy Protection

For real-time recognition only: Voice content is used only for the current recognition and processing session.
No long-term retention: Voice data is not kept long-term.
Not used for model training: Voice content is never used to train any model.

4. System Permission Principles

Permissions serve functions: Used for reading selected text, screenshots, voice input, etc.
Not unlimited collection: Granting permission does not mean continuous monitoring.
On-demand activation: Permissions only take effect when you use the corresponding feature.

5. Privacy Protection Summary

Local-first storage for personal settings and history
Minimal necessary use — data only used for the current function
Voice is not retained — only for real-time recognition
No participation in training — voice data not used for model training
Perceivable permissions — users know what permissions are needed and why

6. User Recommendations

Only enable necessary permissions
Check content before sharing conversations or screenshots
Periodically clean up local records
Allow time for re-setup when changing devices with many local configurations

7. One-Sentence Summary

Owlfy’s privacy policy is not “collect as much as possible,” but:

Store locally when possible

Use only for the current function

Voice is for recognition only — not saved, not trained

Stamina System

Owlfy uses a stamina system to manage AI feature usage quotas.

Item	Description
Stamina Usage	Used for AI-related operations
Daily Stamina	50 stamina gifted daily
Stamina Acquisition	Earn extra stamina through referrals, etc.

Getting Stamina

Method	Reward
Daily check-in	50 stamina/day
Invite friends	1000 stamina/person
Promote Owlfy	Up to lifetime membership

Technical Support

Item	Description
Supported Platforms	Windows, macOS (Intel/Apple Silicon), Linux
Developer Tools	Built-in DevTools for debugging
MCP Ecosystem	Supports installing third-party MCP skills

FAQ

How to customize app aliases?

Find “App Management” in Settings to add custom aliases for apps.

How to add custom websites?

Find “Website Management” in Settings to add URLs and aliases.

How to create custom roles?

Click “Role Management” in the voice input interface to add exclusive roles.

What to do when stamina runs out?

Wait for the next day’s gifted stamina, or earn extra stamina by inviting friends.

Version Information

Current Version: 2.0
Release Date: 2026
Developer: Silicon Geek Co., Limited