GitJarvis

Top 100 trending GitHub AI and ML projects of the week — auto-cloned and updated every week.

GitJarvis runs every Saturday and automatically:

1. Fetches the top 100 trending GitHub AI and ML projects of the week
2. Clones & builds each project (Node, Python, Rust, Go, Docker...)
3. Deploys everything to a live server
4. Serves a live dashboard where anyone can browse and launch them instantly

Live Dashboard: https://gitjarvis.com

---

How It Works

Every Saturday @ 8am UTC
        │
        ▼
  [1] FETCH ──────── GitHub Trending + API → top 100 AI/ML repos
        │
        ▼
  [2] INSTALL ─────── git clone → detect type → npm/pip/uv install → build
        │
        ▼
  [3] DEPLOY ──────── SSH to VPS → upload → PM2 start → nginx route
        │
        ▼
  [4] DASHBOARD ───── Live web UI showing all projects with categories,
                      OS compatibility, prerequisites, and deploy links

---

Quick Start

# Clone
git clone https://github.com/indianbill007/gitjarvis.git
cd gitjarvis

# Install deps
npm install

# Copy env
cp .env.example .env
# Add your GITHUB_TOKEN to .env (optional but avoids rate limits)

# Run full weekly pipeline
node gitjarvis.js weekly

# Or step by step:
node gitjarvis.js fetch           # Fetch trending AI/ML repos
node gitjarvis.js install         # Clone & build all
node gitjarvis.js deploy          # Push to VPS
node gitjarvis.js status          # Check what's running

---

Commands

Command	Description
gitjarvis fetch	Fetch top 100 trending AI/ML repos from GitHub
gitjarvis install	Clone & install all fetched repos
gitjarvis deploy	Deploy to VPS
gitjarvis weekly	Full pipeline (fetch → install → deploy)
gitjarvis setup-server	First-time VPS setup
gitjarvis status	Show current deployment status

---

Dashboard Features

• Categorized view — AI Agents, LLM & Models, AI Coding, RAG & Knowledge, Image & Video AI, Voice & Speech, AI Frameworks & Tools, AI Automation
• OS compatibility icons — Windows, Mac, Linux support indicators per project
• Prerequisites — shows what you need to run each project (e.g. Claude, Chrome, Python, CUDA)
• Deploy button — links to the latest GitHub release when available
• Search & filter — by name, language, status, or category

---

Server

IP	163.227.174.222
Specs	4 core · 8GB RAM · 100GB HDD
OS	Linux (Ubuntu)
Process manager	PM2
Web server	Nginx
Dashboard port	:80

---

Project Structure

gitjarvis/
├── gitjarvis.js          ← CLI entry point
├── src/
│   ├── fetcher.js        ← GitHub trending scraper + API
│   ├── categorizer.js    ← AI filtering & category assignment
│   ├── project-meta.js   ← OS compatibility & prerequisites data
│   ├── release-checker.js← GitHub release fetcher
│   ├── installer.js      ← Clone, detect type, build
│   └── deployer.js       ← SSH deploy to VPS (node-ssh)
├── dashboard/
│   ├── server.js         ← Express API server
│   └── public/
│       └── index.html    ← Web dashboard UI
├── data/                 ← Generated: repos.json, install-results.json
├── projects/             ← Generated: cloned repos
├── .github/workflows/
│   └── weekly.yml        ← GitHub Actions weekly cron
└── .env.example

---

Supported Project Types

Type	Install	Build
Node.js	npm install / bun install	npm run build
Python	pip install / uv sync	—
Docker	docker compose pull	docker compose up
Rust	—	cargo build (if available)
Go	—	go build (if available)

---

Automation

GitJarvis runs automatically via GitHub Actions every Saturday at 8am UTC. You can also trigger it manually from the Actions tab.

Secrets needed in your GitHub repo:

• GITHUB_TOKEN — auto-provided by GitHub Actions
• VPS_HOST, VPS_USER, VPS_PASSWORD — your server credentials

---

License

AGPL-3.0