MetaForge

Integrated software for data transformation, integration, and visualization

What is MetaForge?

MetaForge is integrated software for data transformation, integration, and visualization. It supports a wide variety of data sources — relational databases, NoSQL databases, big data engines, message queues and file systems — and lets you manage multiple data sources, query and transform data, build datasets, and create charts and monitoring views.

Features

Data source management — connect, test and manage 50+ data sources from a single console
SQL query console — IDE-style editor with syntax highlighting, formatting, execution history and export
Datasets — build reusable datasets from tables or ad-hoc queries, preview data and define column metadata
Visualization — drag-and-drop dashboards with tables, pies and more chart types
Reports — schedule and share reports
Workflows — chain data jobs into repeatable workflows
File storage — read and write files through Aliyun OSS, Amazon S3, HDFS, local storage and more
Notifications — DingTalk, email and other channels for job and system events
AI assistant — chat with your data through configured AI providers
Plugin architecture — every connector, converter and storage backend is a hot-loadable plugin
Administration — users, roles, menus and runtime configuration with fine-grained permissions

Requirements

Dependency	Version
JDK	21
Maven	3.9+ (wrapper `./mvnw` included)
Node.js	20
pnpm	9.13.2

Getting Started

Must-read for users: be sure to execute the following command after cloning the code locally

cp configure/git-hook/* .git/hooks
chmod 700 .git/hooks/*

Run the backend (development)

Run the main class org.devlive.metaforge.server.MetaForge (module core/metaforge-server) from your IDE. The server starts on http://localhost:9096. On first launch the setup wizard guides you through selecting the metadata database (embedded H2 by default, or MySQL / PostgreSQL / SQL Server / Oracle) and creating the administrator account.

Run the frontend (development)

pnpm install
pnpm --filter metaforge-web dev

The web console starts on http://localhost:3001 and proxies API calls to the backend.

Build for production

# Frontend
pnpm --filter metaforge-web build

# Backend + distribution package (the server serves the built frontend
# from core/metaforge-web/dist; the tarball lands in dist/)
./mvnw -T1C clean package -DskipTests

Docker

docker pull devliveorg/metaforge:latest
docker run -d --name metaforge -p 9096:9096 devliveorg/metaforge:latest

Or unpack the distribution built locally (dist/metaforge-server-2026.0.0-bin.tar.gz) and run bin/startup.sh.

Project Structure

metaforge
├── client                        # Command line client
├── configure                     # Default configuration and plugin manifests
├── core
│   ├── metaforge-captcha         # Captcha generation
│   ├── metaforge-common          # Shared utilities and enums
│   ├── metaforge-condor          # Scheduling integration
│   ├── metaforge-server          # Spring Boot server (REST API entry point)
│   ├── metaforge-service         # Domain services and persistence
│   ├── metaforge-security        # Security configuration
│   ├── metaforge-sql             # SQL statement builders
│   ├── metaforge-spi             # Plugin SPI contracts
│   ├── metaforge-web             # Vue 3 + Tailwind web console
├── convert                       # Data converters (JSON, CSV, XML, ...)
├── driver                        # Database drivers (Redis, MongoDB, ...)
├── executor                      # Execution engines
├── fs                            # File storage plugins (OSS, S3, local, ...)
├── lib                           # Shared libraries (HTTP client, logging, shell)
├── notify                        # Notification channels (DingTalk, ...)
├── parser                        # SQL parsers (MySQL, Trino, ...)
├── plugin                        # Data source plugins (50+ connectors)
├── scheduler                     # Job scheduling
└── shaded                        # Shaded third-party integrations