User Guide

Everything you need to get started with CrackingInterview.

Contents

Getting Started
Opening Chrome
Selecting an Input Source
Solving — Text Extraction
Solving — Screenshot
Solving — Audio Recording (Pro)
Reading AI Responses
System Design Mode
Global Hotkeys
Custom Prompts
Settings
Free vs Pro
Using Your Own API Key (Free users)
Troubleshooting

1. Getting Started

Prerequisites

Google Chrome must be installed on your computer. The app connects to Chrome to read tabs and extract content.

macOS Permissions

On first launch, macOS will ask you to grant permissions. Go to System Settings → Privacy & Security and enable:

Accessibility — Required for global hotkeys to work system-wide.
Screen & System Audio Recording — Required for screenshot capture, display capture, and audio recording.

Windows Permissions

On Windows, allow the app through Windows Security / SmartScreen when prompted on first launch. No additional system permissions are required.

First Steps

1 Create an account — Open the app and sign up with your email and password.

2 Sign in — Enter your credentials. Check "Remember me" to pre-fill your password next time.

3 Open Chrome — Click "Open Chrome" in the header bar. This launches a Chrome window connected to the app.

4 Navigate to a problem — In the Chrome window, go to a coding problem page.

5 Select your input source — Use the Input Source dropdown to pick the Chrome tab.

6 Hit Solve! — Click the Solve button and wait for the AI response.

Always on Top & Stealth Mode

The app window always stays on top of other windows so your AI assistant is always visible. It also supports stealth mode — hidden from screen-sharing apps (Zoom, Teams), screenshots, Dock (macOS), and Taskbar (Windows). Enable it in Settings → App.

Create Account

2. Opening Chrome

CrackingInterview connects to Google Chrome through a special debugging protocol. Chrome must be installed — just click the "Open Chrome" button in the header.

The app opens a separate Chrome window (it won't interfere with your regular Chrome sessions).
Once connected, the header shows "Chrome Ready" with a green indicator.
Navigate to any website in that Chrome window.

CrackingInterview main screen with Chrome Ready status and LeetCode tab

Main screen — Chrome Ready, LeetCode tab selected

Tip: After Chrome opens and loads a page, click the refresh button next to the Input Source dropdown to update the list of available tabs.

3. Selecting an Input Source

The Input Source dropdown shows all available sources:

Chrome Tabs — Any open tab in the connected Chrome window (shows title + live thumbnail)
Displays — Your screen(s) for full-screen capture (Pro only)
Audio (System) — Record system audio and microphone (Pro only)

Select the source you want to analyze, then click Solve.

Input source dropdown showing Chrome tabs and displays with thumbnails

Input source dropdown — Chrome tabs with thumbnails, displays

Tip: If you opened new tabs or navigated to a new page, press the refresh button to update the list.

4. Solving — Text Extraction

This is the default mode. The app extracts the text content from the selected Chrome tab and sends it to AI.

1 Select a Chrome tab from the Input Source dropdown.

2 Make sure "Text Extraction" is selected in Settings → AI Models → Input Mode.

3 Click Solve (or press Cmd+1).

The progress stepper shows: Extract → Asking AI, along with the AI model and prompt being used. Once complete, the explanation and solution appear.

Progress stepper showing Extract → Asking AI with Gemini 3 Flash model

Progress stepper — shows model (Gemini 3 Flash) and prompt (Algorithm - Optimal)

5. Solving — Screenshot

Screenshot mode captures a visual image of the tab or display and sends it to AI. Useful when the problem includes images, diagrams, or complex formatting.

1 Select your input source (Chrome tab or Display).

2 Select "Screenshot Capture" in Settings → AI Models → Input Mode — or press Cmd+2.

3 Click Solve (or press Cmd+2).

AI Models settings showing Input Mode — Text Extraction and Screenshot Capture toggle

Settings → AI Models — Input Mode toggle

Tip: If your input source is a Display, the app always uses screenshot mode automatically, even if Text Extraction is selected.

6. Solving — Audio Recording (Pro)

Audio mode records system audio (e.g., from a Zoom or Teams call) and sends the recording to AI for analysis. Designed for verbal interview practice. Supports 20+ languages with auto-detection.

1 Select "Audio (System)" from the Input Source dropdown.

2 The app automatically selects the "Verbal Interview" prompt.

3 Choose your interview language (or leave on Auto-detect).

4 Click Record (or press Cmd+3) to start recording.

5 The timer shows recording duration. Speak or let the audio play.

6 Click Stop (or press Cmd+3 again) to stop and send to AI.

Animated demo of audio recording — system audio capture, transcription, and AI response

Audio recording flow — record, transcribe, and get AI response

Audio language selection dropdown with 20+ languages including auto-detect

Interview language selection — 20+ languages supported

Note: Audio recording has a 3-minute maximum per session. The recording automatically stops and sends to AI after 3 minutes. Audio is a Pro-only feature with up to 10 hours per month.

7. Reading AI Responses

After the AI processes your request, you'll see two sections:

Explanation — A concise walkthrough of the approach, algorithm, and complexity analysis. Click the header to collapse/expand.
Solution — The complete code solution with syntax highlighting. Click "Copy Code" to copy it to your clipboard.

Both sections can be closed with the ✕ button. They automatically reappear when you run a new Solve.

AI explanation with approach, complexity analysis, and syntax-highlighted code solution

Explanation + Solution with syntax-highlighted code and Copy Code button

8. System Design Mode

The System Design prompt generates comprehensive designs with rendered diagrams — not just text. Select "System Design" in Settings → Prompts, then Solve using a screenshot of the design problem.

The AI returns:

Explanation — Functional & non-functional requirements, key trade-offs
Design — Rendered architecture diagrams, ER data models, sequence flows, and a scaling strategy

Animated demo of system design — architecture diagrams, ER models, and sequence flows rendered in real time

System design flow — architecture, ER model, sequence diagram, and scaling strategy

9. Global Hotkeys

Hotkeys work even when the app is in the background, so you can trigger actions without switching windows.

Default Shortcut	Action
Cmd+1	Extract text → Solve
Cmd+2	Screenshot → Solve
Cmd+3	Start/Stop audio recording → Solve
Cmd+Up / Down	Scroll explanation up/down
Cmd+Shift+Arrow	Move the app window
Cmd+Shift+H	Show/Hide the app window
Cmd+Shift+Q	Quit the app

All hotkeys are customizable in Settings → HotKeys. On Windows, Cmd is replaced with Alt.

HotKeys settings — Solve, Navigation, and App sections with customizable shortcuts

Settings → HotKeys

Tip: Use Cmd+Shift+H to quickly hide/show the app while working in Chrome. Hotkeys work globally — no need to switch windows.

10. Custom Prompts

Prompts control how the AI analyzes your problem. The app comes with 6 built-in prompts:

Algorithm - Optimal — Finds the most efficient solution
Algorithm - Beginner — Explains in simpler terms
System Design — Generates architecture diagrams, data models, and scaling plans
Code Review — Reviews and improves existing code
Explain Concept — Explains algorithms and concepts
Verbal Interview — For audio-based interview practice

Editing a Prompt

Go to Settings → Prompts and click "Edit" on any prompt. You can customize the system instructions and user template. Use {CONTENT} as a placeholder for the extracted text and {LANGUAGE} for the programming language.

Creating a Custom Prompt

Click "+ New Prompt" to create your own. You can have up to 3 custom prompts.

Programming Language

Select your preferred language in Settings → Prompts. Available: Java, Python, JavaScript, C++, Swift, Go, PHP, Ruby, SQL.

Settings → Prompts

Prompt editor with system prompt and user prompt fields

Editing a prompt — System & User Prompt

11. Settings

Click the gear icon in the header to open Settings. Tabs include:

Account — Subscription status, AI call and audio usage quotas, manage billing, sign out
AI Models — Choose which AI model to use (Gemini 3 Flash, GPT-5, Claude Sonnet, Grok). Switch between Text Extraction and Screenshot Capture input modes. Free users can enter their own API key
Prompts — Select, edit, create, and manage prompts. Choose your programming language
HotKeys — Customize all keyboard shortcuts
App — Adjust window transparency, switch between Light/Dark theme, and toggle Stealth Mode

AI Models

Account

App Settings

12. Free vs Pro

Feature	Free	Pro ($15/mo)
AI calls	3 calls (or unlimited with your own API key)	150 calls per month
AI models	1 model (or your own via API key)	All 4 premium models
Audio recording	✗	10 hours per month
Input sources	Chrome tabs only	Chrome tabs, Display, Audio
Websites	Coding practice sites only	Any website
Text Extraction	✓	✓
Screenshot Capture	✓ (Chrome tabs)	✓ (Chrome tabs + Display)
System Design Diagrams	✓	✓
Custom Prompts	✓	✓
Global Hotkeys	✓	✓
Always on Top	✓	✓
Stealth Mode	✓	✓
Light / Dark Theme	✓	✓
Window Transparency	✓	✓

To upgrade, go to Settings → Account and click "Manage Subscription". See how we compare to other AI interview tools.

13. Using Your Own API Key (Free Users)

Free users can bring their own API key to get unlimited AI calls:

1 Go to Settings → AI Models.

2 Enter your API key in the provided field.

3 Click Save. The app will use your key for all AI requests.

Note: Your API key is stored locally on your device only. We never send it to our servers. You are responsible for any charges from the API provider.

14. Troubleshooting

"Open Chrome" button doesn't connect

Make sure Google Chrome is installed on your computer.
Make sure no other application is using debugging port 9222.
Try closing all Chrome windows and clicking "Open Chrome" again.

Screenshot or audio capture not working (macOS)

Go to System Settings → Privacy & Security → Screen & System Audio Recording and enable CrackingInterview.
You may need to restart the app after granting this permission.
If it still doesn't work, try toggling the permission off and back on.

Input Source dropdown is empty

Click the refresh button next to the dropdown.
Make sure Chrome is connected (header shows "Chrome Ready").
Open at least one tab in the Chrome window.

AI response is slow or fails

Check your internet connection.
If using your own API key, verify it's valid and has available quota.
Try selecting a different AI model in Settings → AI Models.
Complex problems (especially System Design with screenshots) may take longer — the app has a 50-second timeout.

"Domain not allowed" error (Free users)

Free tier is limited to supported coding practice websites.
Navigate to a supported site in the Chrome window.
Upgrade to Pro for access to any website.

Audio recording doesn't work

Audio is a Pro-only feature. Make sure you have an active Pro subscription.
On macOS, grant Screen & System Audio Recording permission in System Settings.
Make sure system audio is playing (the app records what your computer outputs).
If audio worked once but stopped, try toggling the Screen Recording permission off and back on in System Settings.

Hotkeys don't work

On macOS, go to System Settings → Privacy & Security → Accessibility and enable CrackingInterview.
You may need to restart the app after granting this permission.
Check for conflicts with other apps using the same shortcuts.

Still need help? Contact us at [email protected]