AppForgeHub

AppForgeHub

Docker app from allornothing's Repository

Overview

Self-hosted AI app builder: describe an app in chat and it scaffolds a real, running web app (Express + SQLite + built-in auth) inside its own sandboxed process, with a live preview and an editable file tree -- no cloud account, billing, or account creation involved. AI is fully pluggable and optional. Point it at a cloud API (Anthropic, OpenAI, or any OpenAI-compatible endpoint) for the best results, or run entirely offline against a local Ollama or llama.cpp server -- switch anytime from Settings. Nothing is generated until you configure a provider. First visit creates the single admin account for this instance (no other signup exists). Generated apps run as child processes inside this container on internal ports 4100-4999 and are reverse-proxied through the same web UI port, so nothing extra needs to be exposed.

AppForge Hub on Unraid

Self-hosted AI app builder: describe an app in chat, get a real running web app with its own database and auth already wired up. AI is optional and pluggable — a cloud API key or a local Ollama/llama.cpp instance.

Setup

  1. Apps tab → search "AppForge Hub" → Install (or add the template manually — see the parent repo's README).
  2. Set the Appdata path (default /mnt/user/appdata/appforge-hub) — this holds the admin account, settings, and every generated app's files and database. Back this up like you would any other app's config.
  3. Start the container and open the WebUI. The first visit creates the single admin account for this instance — there is no other signup.
  4. (Optional) In Settings, choose an AI provider:
    • Anthropic / OpenAI / OpenAI-compatible — paste an API key. Use "OpenAI-compatible" to point at OpenRouter, Groq, or any other endpoint that speaks the OpenAI API shape.
    • Ollama / llama.cpp — point the base URL at an existing instance on your LAN (e.g. http://192.168.1.50:11434), or run one alongside this container (see below). Nothing leaves your network.

Until a provider is configured, the dashboard and project pages work, but chat-based generation is disabled.

Running a local model alongside it

Easiest path: install the Ollama template from Community Applications separately, pull a model (docker exec -it ollama ollama pull llama3.1), then in AppForge Hub's Settings set provider = Ollama and base URL to http://<unraid-ip>:11434.

Notes

  • Generated apps run as child Node processes inside this container on internal ports 4100–4999 and are reverse-proxied through the same WebUI port — nothing else needs to be exposed or port-forwarded.
  • Set AI_PROVIDER / AI_API_KEY / AI_BASE_URL / AI_MODEL as container variables to pre-configure a provider at first boot; the Settings page can still override them afterward (values saved there take precedence).
  • All generated-app data lives under /config/projects/<id>/ inside the Appdata path, including each app's own data.sqlite — deleting an app from the UI deletes its folder.

Install AppForgeHub on Unraid in a few clicks.

Find AppForgeHub in Community Apps on your Unraid server, review the template, and click Install. Unraid handles the Docker app or plugin setup from the published template.

Open the Apps tab on your Unraid server Search Community Apps for AppForgeHub Review the template variables and paths Click Install

Download Statistics

101
Total Downloads

Related apps

Details

Repository
allornothing/appforge-hub:latest
Last Updated2026-09-02
First Seen2026-09-03

Runtime arguments

Web UI
http://[IP]:[PORT:8000]/
Network
bridge
Privileged
false

Template configuration

WebUI PortPorttcp

Web UI / API port

Target
8000
Default
8000
Value
8000
AppdataPathrw

Admin account, settings, and every generated app's files + database

Target
/config
Default
/mnt/user/appdata/appforge-hub
Value
/mnt/user/appdata/appforge-hub
PUIDVariable

Unraid's nobody user -- keeps appdata editable from the host

Default
99
Value
99
PGIDVariable

Unraid's users group

Default
100
Value
100
TZVariable

Container timezone

Default
Etc/UTC
Value
Etc/UTC
AI ProviderVariable

Optional, can also be set from the Settings page after install: none | anthropic | openai | openai_compatible | ollama | llamacpp

Target
AI_PROVIDER
Default
none
Value
none
AI API KeyVariable

Used when AI Provider = anthropic, openai, or openai_compatible

Target
AI_API_KEY
AI Base URLVariable

Used when AI Provider = openai_compatible, ollama, or llamacpp -- e.g. http://ollama:11434 or http://llamacpp:8080

Target
AI_BASE_URL
AI ModelVariable

Model name/ID for the selected provider

Target
AI_MODEL