- ▸GitHub Repository: https://github.com/SadeepaNHerath/MetaScribe
- ▸Website: https://meta-scribe.vercel.app/
Overview
MetaScribe is a full-stack SEO analysis platform that helps developers, marketers, and site owners evaluate how well a webpage is optimized for search and social sharing.
Given a URL, it fetches the page, parses its HTML, and generates a multi-dimensional SEO report including:
- ▸Meta tag coverage
- ▸Social sharing tags
- ▸Best-practice checks
- ▸Structured data validation
- ▸Content quality analysis
The result is a practical, score-based report with concrete recommendations and copy-ready examples.
The Problem
Many SEO tools are either too shallow (basic tag checkers) or too heavy (enterprise dashboards with noisy output). MetaScribe focuses on a useful middle ground: clear diagnostics, transparent scoring, and actionable fixes for technical SEO essentials.
Key Features
- ▸URL-based SEO scan with automatic protocol handling and robust fetch error messaging.
- ▸Multi-category scoring system:
- ▸Overall score
- ▸Required tags
- ▸Social tags
- ▸Best practices
- ▸Content quality
- ▸Structured data
- ▸Meta tag extraction across:
- ▸Essential tags (title, description, viewport)
- ▸Open Graph and Twitter card tags
- ▸Canonical, robots, language, and additional extended tags
- ▸Structured data analyzer:
- ▸Detects JSON-LD blocks
- ▸Validates schema basics and common schema types
- ▸Flags missing required fields
- ▸Content quality analyzer:
- ▸H1 and heading hierarchy checks
- ▸Image alt-text coverage
- ▸Word count depth analysis
- ▸Platform preview simulation for:
- ▸X/Twitter
- ▸Recommendation engine with severity levels and optional code snippets.
- ▸Raw head HTML inspection and copy support.
- ▸Caching layer for recent analyses with a one-hour TTL.
- ▸Dual runtime model:
- ▸Local Express API
- ▸Serverless API for Vercel deployment
Technical Architecture
MetaScribe uses a React frontend with a Node/Express API for local development, plus a serverless function path for production deployment.
Core SEO analysis logic is centralized in shared modules so extraction, scoring, and recommendations remain consistent across environments.
Stack
| Layer | Technology |
|---|---|
| Frontend | React + TypeScript + Vite |
| UI System | Tailwind CSS + Radix UI + shadcn/ui |
| Data Fetching | TanStack Query |
| Backend (Local) | Node.js + Express |
| Backend (Deploy) | Vercel Serverless Functions |
| HTML Parsing | Cheerio |
| HTTP Client | Axios |
| Validation | Zod |
| Shared Core | TypeScript shared analyzers |
| Testing | Vitest |
| Data Layer | In-memory cache/storage (Drizzle/Postgres planned) |
Status
MetaScribe is under active development with working end-to-end analysis, scoring, previews, and recommendations.
It currently uses in-memory storage, with persistent database integration planned.