LLMScore

Get your website's LLM friendliness score by analyzing its structure and content.

Features

• User Authentication: Secure login system with user-specific data isolation
• Website Mapping: Uses Firecrawl to map all URLs on your website
• LLM Score Evaluation: AI-powered scoring using OpenAI analysis
• Search Performance Analysis: Real search ranking analysis using Firecrawl search API
• SEO Optimization Analysis: Identifies HTML pages missing titles or meta descriptions (excludes static assets)
• AI Optimization Check: Checks for AI-friendly files (/agents.txt, /agent.txt, /llm.txt, /llms.txt)
• Persistent Storage: All evaluations saved to Convex database per user
• Dashboard: View your analysis history and track improvements
• Clean UI: Simple black and white design with monospace font
• Fast Analysis: Quickly get all links and structure of your website

Setup

1. Install dependencies:

   npm install

2. Get API keys:

   • Firecrawl: Sign up at Firecrawl.dev and get your API key
   • OpenAI: Sign up at OpenAI Platform and get your API key
   • Google OAuth: Create credentials at Google Cloud Console
   • Convex: Sign up at Convex.dev for database persistence

3. Configure environment:

   • Copy .env.local and add your API keys:

   FIRECRAWL_API_KEY=your_firecrawl_api_key_here
   OPENAI_API_KEY=your_openai_api_key_here
   NEXT_PUBLIC_CONVEX_URL=your_convex_url_here
   GOOGLE_CLIENT_ID=your_google_client_id_here
   GOOGLE_CLIENT_SECRET=your_google_client_secret_here
   BETTER_AUTH_SECRET=your_secure_random_string_here
   

4. Set up Convex database:

   npx convex dev

   Follow the prompts to create/connect to a Convex project, then deploy the schema:

   npx convex deploy

5. Run the development server:

   npm run dev

6. Open http://localhost:3000 in your browser

Usage

1. Sign In: Click the "Sign In" button to authenticate
2. Enter Website: Enter your website domain (e.g., example.com - no need for https://)
3. Analyze: Click "Analyze Website" to run comprehensive evaluation
4. View Results: See your LLM Score with detailed breakdown:
   • Overall score (1-10 scale)
   • Search visibility analysis
   • Content quality assessment
   • Technical SEO evaluation
   • AI optimization status
5. Check History: Visit /dashboard to view all your past analyses
6. Track Progress: Monitor improvements over time with persistent storage

How it works

The app provides user-authenticated, comprehensive LLM compatibility analysis:

Website Mapping & SEO Analysis

Uses Firecrawl's /map endpoint to:

• Crawl your website and find all URLs
• Include sitemap data when available
• Return up to 100 links for analysis
• Analyze SEO metadata: Identify HTML pages missing titles or meta descriptions (excludes static assets like .css, .js, images, etc.)

AI Optimization Files Check

Checks for AI-friendly configuration files:

• /agents.txt - Instructions for AI agents
• /agent.txt - Alternative agent instructions
• /llm.txt - LLM-specific instructions
• /llms.txt - Multiple LLM instructions

LLM Score Evaluation

Uses OpenAI's advanced AI models to evaluate:

• Search Visibility: How well the site appears in AI-powered search results
• Content Quality: Structure and metadata completeness
• Technical SEO: Overall technical foundation
• AI Optimization: Presence of AI-friendly configurations

Search Performance Analysis

Uses Firecrawl's search API to perform real-world search analysis:

• Content-Based Keywords: AI analyzes actual website content to generate relevant search terms
• Website Scraping: Scrapes the landing page content for accurate keyword generation
• Search Rankings: Checks actual search result positions for each keyword
• Performance Metrics: Appearance rate, top 10 rankings, average position
• Detailed Insights: Shows where the site ranks for specific search queries

The evaluation provides a weighted overall score and actionable recommendations for improving LLM compatibility and search visibility.

User Data Management

• Authentication Required: All analyses require user login
• Data Isolation: Each user's evaluations are stored separately
• Persistent Storage: Results saved to Convex database for long-term access
• History Tracking: View and compare past analyses in the dashboard
• Secure Access: User-specific data protection and privacy

Next Steps

Future enhancements could include:

• LLM scoring algorithm based on content analysis
• SEO optimization suggestions
• Accessibility checks
• Performance metrics