Skip to content

setup.md

Latest commit

 

History

History
96 lines (69 loc) · 3.46 KB

setup.md

File metadata and controls

96 lines (69 loc) · 3.46 KB

Setup Guide

📦 Installation

  1. Clone the repository: 
    git clone https://github.com/hellosaumil/JobPostHighlightsExtension.git
  2. Open Chrome and navigate to chrome://extensions/
  3. Enable Developer mode (toggle in top right)
  4. Click Load unpacked and select the extension folder
  5. Pin the extension to the toolbar for quick access

⚙️ Provider Configuration

Open the extension → click the Settings (⚙️) icon → choose your AI provider.

On-Device (Gemini Nano)

Tip

Free, private, and fast — no API key needed. Recommended for Stage 1 pre-extraction.

Requirements:

  • Chrome 127+ (Dev/Canary channel recommended)
  • ~1.5 GB free disk space for the model download

Setup:

  1. Navigate to chrome://flags
  2. Enable #prompt-api-for-gemini-nano → Set to Enabled
  3. Enable #optimization-guide-on-device-model → Set to Enabled BypassPerfRequirement
  4. Restart Chrome
  5. Check chrome://components → Look for Optimization Guide On Device Model
    • If version shows 0.0.0.0, click Check for update
    • Wait for the model to download (~1.5 GB)

Note

If you see "not enough space" errors, the model needs ~1.5 GB of free disk space. Free up space or switch to a different provider.

Troubleshooting:

Gemini Cloud API

  1. Get a free API key from Google AI Studio
  2. In Settings, select Google Gemini (Cloud) as the provider
  3. Choose your model:
    • Gemini 3.1 Pro Preview — Best quality
    • Gemini 3.1 Flash-Lite Preview — Faster, lower cost
  4. Paste your API key → Click Save Settings

Important

Your API key is stored locally in chrome.storage.local and never leaves your browser.

Ollama (Local)

  1. Install Ollama from ollama.com
  2. Pull a model: 
    ollama pull gemma3:4b    # Recommended for quality/speed balance
  3. In Settings, select Ollama (Local) as the provider
  4. Set the server URL (default: http://localhost:11434)
  5. Click the refresh button to load available models
  6. Select your model → Click Save Settings

See Ollama Guide for CORS troubleshooting and performance tips.

📄 Resume Configuration

The candidate resume is embedded as a text summary inside prompts/stage_2.md (substituted into the {{resumeSource}} placeholder at runtime by fetchPrompt() in ai_service.js).

To customize for your own resume, edit the hardcoded summary string in ai_service.js inside fetchPrompt():

// In ai_service.js → fetchPrompt()
const resumeSummary = "Your Name (Role, X years exp, Key Skills).";

Note: An optional resume.pdf path exists in the codebase for future Gemini Cloud file-upload support, but is not currently used.

⌨️ Keyboard Shortcut

Platform Shortcut Action
Mac Cmd+J Open side panel
Windows/Linux Ctrl+J Open side panel

Customize in chrome://extensions/shortcuts.