AI Intelligence Hub

Track AI trends from 19 sources in one dashboard. GitHub Trending, HuggingFace, MCP Servers, Claude Code Releases, Anthropic Skills & Cookbooks, arXiv, Google AI Blog, Simon Willison, and more — with full-text search and keyword scoring.

]() ]()

Quick Start

git clone https://github.com/ytrofr/ai-intelligence-hub.git
cd ai-intelligence-hub
npm install
cp config/projects.example.json config/projects.json   # then edit with your own projects
node server.js
# Open http://localhost:4444

config/projects.json is gitignored — it holds your personal project portfolio. The repo ships config/projects.example.json as a template; copy it and customize the projects array. The claude-ecosystem entry works as-is.

No external services required. All 19 sources use free, unauthenticated APIs. Fetches all sources in parallel (~5 seconds).

Optional: Add a GITHUB_TOKEN in .env to increase GitHub API rate limit from 60/h to 5,000/h. See .env.example.

Features

• 19 AI Sources — GitHub Trending repos, HuggingFace models, MCP server registry, Claude Code releases and docs, Anthropic Skills Library, Claude Cookbooks, arXiv CS.AI, Google AI Blog, Simon Willison, MarkTechPost, The Gradient, Hacker News, Product Hunt, Anthropic Blog, OpenAI Blog, TechCrunch AI, MIT AI News, AI News
• Full-Text Search — SQLite FTS5 indexes all items for instant keyword search
• Keyword Scoring — Configurable categories with weighted keywords rank items by relevance
• Bookmarks — Save items for later with persistent bookmarks
• Modular Architecture — Add new sources by extending BaseModule (one file per source)
• Zero Build Step — Vanilla JavaScript frontend, no bundler or framework required
• Self-Contained — SQLite database, no external database or service dependencies

Sources

Architecture

server.js (port 4444)
├── modules/           # Source fetchers (BaseModule pattern)
│   ├── base-module.js # Abstract base with normalize()
│   ├── github.js      # GitHub trending repos
│   ├── huggingface.js # HuggingFace models
│   ├── rss.js         # RSS/Atom feeds (12 sources)
│   ├── mcp-registry.js# MCP server registry
│   └── changelog.js   # Claude Code releases + docs
├── config/
│   ├── sources.json   # Source definitions (URL, type, refresh)
│   └── keywords.json  # Scoring categories and weights
├── database/          # SQLite with FTS5 full-text search
│   ├── db.js          # Database operations
│   └── schema.sql     # Table definitions
├── routes/            # Express API routes
│   ├── items.js       # List/filter items
│   ├── fetch.js       # Trigger source fetches
│   ├── sources.js     # Source management
│   ├── bookmarks.js   # Bookmark CRUD
│   ├── stats.js       # Dashboard statistics
│   └── search.js      # FTS5 search + suggestions
└── public/            # Frontend (vanilla JS)
    ├── index.html     # Single-page dashboard
    ├── css/           # Styles (variables, layout, components)
    └── js/            # App logic (api, filters, ui, icons)

API Reference

Adding a New Source

1. Create a module in modules/ extending BaseModule
2. Implement the fetch() method returning normalized items
3. Register it in modules/index.js
4. Add source config to config/sources.json
5. Add badge CSS in public/css/components.css

See CONTRIBUTING.md for details.

Configuration

Keyword Scoring

Edit config/keywords.json to customize scoring categories:

{
  "categories": [
    {
      "id": "mcp",
      "label": "MCP & Tools",
      "keywords": ["mcp", "model context protocol", "tool-use"],
      "weight": 3
    }
  ]
}

Source Configuration

Edit config/sources.json to add or modify sources:

{
  "sources": [
    {
      "id": "github-trending",
      "name": "GitHub Trending",
      "type": "github",
      "url": "https://api.github.com/search/repositories",
      "refresh_minutes": 60
    }
  ]
}

Requirements

• Node.js 18+
• No external databases (SQLite is embedded)
• No API keys required (all sources use free APIs)
• Optional: GitHub personal access token for higher rate limits

FAQ

Q: What port does the Intelligence Hub run on? A: Port 4444. This is hardcoded in server.js and does not conflict with other common development ports.

Q: Do I need any API keys? A: No. All 19 sources use free, unauthenticated APIs. Optionally add a GITHUB_TOKEN for higher GitHub rate limits (60/h free, 5,000/h with token).

Q: How is data stored? A: SQLite with FTS5 (full-text search). The database is created automatically on first run. No setup needed.

Q: Can I add my own sources? A: Yes. Create a module extending BaseModule, register it, and add config. See CONTRIBUTING.md.

Q: Is there a hosted version? A: No. This is a local-first tool designed to run on your machine. Your data stays on your machine.

Built With

• Express — Web framework for Node.js
• better-sqlite3 — Synchronous SQLite3 with FTS5 support
• xml2js — XML/RSS feed parser

License

MIT