Fractera

@Roma Bolshiyanov (Armstromg)

Visit Server

13 hours ago

Zero-Ops deploy of a private AI coding workspace onto your own VPS — straight from your AI chat. Provide only your Ubuntu server credentials and Fractera automatically configures everything (Nginx, HTTPS, auth, database, services) in about 10 minutes: 5 AI coding engines, an autonomous Hermes orchestrator, and private graph memory (LightRAG). No terminal, no DevOps. The deployment is IP-first and free; attaching a custom domain with HTTPS is an optional later step. The connector can register the user, recommend a VPS, run and monitor the full deploy, and answer questions about the project via get_project_info.

Overview Tools Comments

Overview

Fractera

Zero-Ops deploy of a private AI workspace to your own VPS — straight from your AI chat.

Fractera — step by step

Fractera turns a bare Ubuntu VPS into a complete, private AI coding workspace — without any DevOps. You give your AI agent only the server credentials, and the connector configures everything automatically (Nginx, HTTPS, auth, database, services) in about 10 minutes.

What you get on your own server

5 AI coding engines — Claude Code, Codex, Gemini CLI, Qwen, Kimi (switchable on the fly)
Hermes — an autonomous orchestrator agent that can work while you sleep
Private graph memory (LightRAG) over your code and docs
Fully self-hosted on your hardware — nothing leaks to third-party clouds

AI chat

How to use

Add this connector to your AI client by URL.
In chat, ask to deploy Fractera and provide your VPS IP + credentials.
The agent runs the full deploy and reports back when it's live at http://<your-ip>:3002.

Tools

register_and_deploy — create the user, register the server, run the full deploy
retry_deploy — recover a failed deployment
check_status — check deployment progress / state
get_vps_recommendation — suggest a suitable VPS
get_subdomain — return the workspace address
get_project_info — full architecture, specs and FAQ about the project

Real-world use cases

Private team content workspace (collaborate, nothing public)
Local-business lead dispatcher with an internal Kanban board
Hybrid public/private adaptive AI tutor controlled via Telegram
Autonomous trend-scraping blog that publishes and reports traffic to Telegram

Good to know

Free — Fractera never charges for the deployment; it runs on your own hardware.
IP-first — finishes on plain HTTP at your server IP; a custom domain with HTTPS is an optional later step inside the workspace.

Full knowledge base → fractera.ai/mcp-info

Try in Playground

Server Config

{
  "mcpServers": {
    "fractera": {
      "url": "https://www.fractera.ai/api/mcp"
    }
  }
}

Build with ShipAny.