.jpg)
Scan. Explain. Document. Ship.
DataForge turns any codebase into structured knowledge — architecture reports, dependency maps, onboarding guides, and production-grade documentation — generated by AI, without ever leaving your shell.
Tip
New to the command line? You don't need Git. Download the repo as a ZIP, extract it, and run one command — see Installation below.
- Why DataForge
- Quick Start
- How DataForge Works
- Installation
- Run DataForge From Anywhere
- Command Reference
- Plans & Licensing
- How Hybrid Licensing Works
- AI Providers
- Project Structure
- Security & Local-First Architecture
- Roadmap
- Contributing
- License
Reading someone else's codebase — or your own from six months ago — takes time you don't have.
DataForge CLI automates that first pass.
Point it at a project. DataForge scans the codebase, builds structured context, and sends that context to your selected AI provider to generate useful developer intelligence.
No copy-pasting files into a chat window. No manually building architecture documentation. No leaving your terminal.
Your code never touches a DataForge server.
Reports are generated locally and saved locally. The only network requests DataForge makes are to your selected AI provider, using your own API key — or to no network at all when using Ollama locally.
Clone the repository:
git clone https://github.com/RabbitGamesDev/DataForge-CLI.git
cd DataForge-CLI/dataforge-cliInstall dependencies:
pip install -r requirements.txtLaunch DataForge:
python main.pyThe first launch opens an interactive setup wizard where you can configure your AI engine, language, and terminal theme.
Then scan your first project:
python main.py scan .Your generated reports will appear in:
dataforge-reports/
That's it. Your codebase just became AI-readable knowledge.
YOUR CODEBASE
│
▼
┌─────────────────────┐
│ DATAFORGE SCANNER │
└──────────┬──────────┘
│
┌────────────────┼────────────────┐
▼ ▼ ▼
Architecture Dependencies Context
│ │ │
└────────────────┼────────────────┘
│
▼
┌─────────────────────┐
│ AI ENGINE │
│ │
│ Groq / OpenAI │
│ Gemini / Claude │
│ Ollama │
└──────────┬──────────┘
│
▼
┌─────────────────────┐
│ STRUCTURED OUTPUT │
└─────────────────────┘
│
┌────────────────┼────────────────┐
▼ ▼ ▼
AI Reports Documentation Architecture
Risk Analysis Onboarding Maps
DataForge is designed to turn a codebase from something you have to understand into something you can immediately work with.
Option 1 — Git Clone (recommended for developers)
git clone https://github.com/RabbitGamesDev/DataForge-CLI.git
cd DataForge-CLI/dataforge-cli
pip install -r requirements.txt
python main.pyOption 2 — Download as ZIP (no Git required)
- Open the DataForge CLI Website Installation.
- Click ZIP → Download DataForge-CLI-main.zip.
- Extract the archive anywhere on your computer.
- Open a terminal inside the extracted
dataforge-clifolder. - Install the dependencies:
pip install -r requirements.txt
- Start DataForge:
python main.py
Option 3 — One-Click Installation
A one-line installer is also available through the official DataForge CLI website for Windows, macOS, and Linux.
On Windows, you can create a launcher so you don't need to navigate into the project directory every time.
Create a file named:
dataforge.bat
Add:
@echo off
python "C:\PATH\TO\DataForge-CLI\dataforge-cli\main.py" %*Add the folder containing the launcher to your system PATH.
You can then run:
dataforge scan .from anywhere.
Core Commands
| Command | Description | Tier |
|---|---|---|
python main.py scan . |
Full project scan → architecture, risks, and suggestions | Free |
python main.py explain main.py |
Explain a specific file | Free |
python main.py map . |
Generate an ASCII architecture / dependency map | Free |
python main.py ask . |
Interactive AI chat with project context | Free |
python main.py onboard . |
Generate a developer onboarding guide | Free |
python main.py welcome |
Open the DataForge welcome letter | Free |
Advanced Commands
| Command | Description | Tier |
|---|---|---|
python main.py doc++ . |
Production-ready documentation | ⭐ Pro / 💜 Teams |
python main.py architecture++ . |
Mermaid architecture diagrams + deep analysis | ⭐ Pro / 💜 Teams |
python main.py export . --html |
Export reports to Markdown, JSON, or HTML | ⭐ Pro / 💜 Teams |
python main.py preset onboarding . |
Run automated multi-step pipelines | ⭐ Pro / 💜 Teams |
Configuration & Licensing
| Command | Description | Tier |
|---|---|---|
python main.py setup |
Configure AI engine, language, and terminal theme | All |
python main.py license YOUR_KEY |
Activate or check a license | All |
python main.py license status |
Check current license status | All |
python main.py help |
Display available commands | All |
Free tier note:
setupuses Groq as the default AI engine and provides the default terminal theme. Pro unlocks additional terminal themes. Teams unlocks full multi-provider selection.
DataForge follows an Open-Core model.
The core CLI remains available under the Apache License 2.0, while Pro and Teams unlock advanced functionality and commercial workflows.
| Feature | 🟢 Free | ⭐ Pro | 💜 Teams |
|---|---|---|---|
scan, explain, map, ask, onboard |
✅ | ✅ | ✅ |
| Default terminal theme | ✅ | ✅ | ✅ |
| Additional terminal themes | — | ✅ | ✅ |
doc++ |
— | ✅ | ✅ |
architecture++ |
— | ✅ | ✅ |
export |
— | ✅ | ✅ |
| Automated presets | — | ✅ | ✅ |
| Multi-provider AI | — | — | ✅ |
| Enterprise report branding | — | — | ✅ |
| Custom report signature | — | — | ✅ |
→ See full pricing on the DataForge website
DataForge uses a hybrid licensing model designed to protect commercial licenses without making the CLI dependent on a permanent internet connection.
┌────────────────────────┐
│ dataforge license │
│ YOUR-KEY │
└────────────┬───────────┘
│
▼
Lemon Squeezy validation
│
▼
┌────────────────────────┐
│ ~/.dataforge/ │
│ license.json │
└────────────┬───────────┘
│
Local validation
│
┌────────────┴────────────┐
│ │
Every 7 days No internet
│ │
▼ ▼
Heartbeat check Offline grace period
│ │
└────────────┬────────────┘
▼
DataForge continues
working locally
- Local-first validation — license information is cached locally for fast routine checks.
- Periodic verification — DataForge periodically verifies the commercial license status.
- Offline grace period — temporary loss of internet connectivity does not immediately interrupt local development.
- No DataForge cloud dependency — DataForge does not require a permanent connection to a DataForge server.
Current status: Lemon Squeezy integration is being finalized for automated checkout, license delivery, and full customer synchronization.
DataForge is built around a multi-provider architecture.
| Provider | Strength | Requires |
|---|---|---|
| Groq | Ultra-fast responses | API key |
| OpenAI | Strong logical reasoning | API key |
| Gemini | Large-context, multi-step analysis | API key |
| Claude | Production-quality technical writing | API key |
| Ollama | Fully local and offline AI | Local Ollama installation |
Configure your provider with:
python main.py setupProvider availability depends on your current DataForge plan.
When using Ollama with a locally installed model, DataForge can operate without sending project data to an external AI provider.
Your Code
│
▼
DataForge CLI
│
▼
Local Ollama Model
│
▼
Generated Report
DataForge-CLI/
├── README.md
├── LICENSE
├── CHANGELOG.md
├── CONTRIBUTING.md
├── SECURITY.md
│
└── dataforge-cli/
├── dataforge/
│ ├── __init__.py
│ ├── api_handler.py # Multi-provider AI routing
│ ├── config_manager.py # config.json + license.json management
│ └── core.py # File scanning, reading, report saving
├── dataforge-reports/ # Generated reports land here
├── .gitignore
├── main.py # CLI entry point + command router
└── requirements.txt
DataForge is designed with a local-first philosophy.
- API keys are stored locally under
~/.dataforge/. - On Windows, the configuration directory is located under
%USERPROFILE%\.dataforge. - License information is stored locally for validation and offline operation.
- Generated reports remain on your machine.
- DataForge itself does not upload your codebase to a DataForge server.
- External AI requests are made to the provider you configure.
- Ollama can be used for fully local AI processing.
- Sensitive local configuration files are excluded through
.gitignore.
Under normal operation, DataForge does not send your project to a DataForge-controlled cloud service.
- When using a cloud AI provider such as Groq, OpenAI, Gemini, or Claude, the relevant project context is sent directly to that provider's API according to your configuration and their applicable policies.
- When using Ollama locally, processing can remain entirely on your machine.
- v1.0 — Interactive setup, Groq integration, and modular report generation
- v1.5 — Multi-provider support, offline Ollama integration, and expanded commands
- v2.0 — Hybrid licensing architecture
- v2.0 — Pro & Teams tiers
- v2.0 — Multi-format export
- v2.0 — Official website
- v2.0 — Customer portal foundation
- v2.0 — Live web dashboard
- v2.0 — Google Analytics integration
- v2.0 — Automatic browser-language translation
- v2.1–v2.3 — Fully automated Lemon Squeezy checkout
- v2.1–v2.3 — Instant license delivery
- v2.1–v2.3 — Real-time customer dashboard synchronization
- v2.5 — Community plugin system
- v2.5 — Integrated static security analysis (SAST)
- v2.5+ — Secure cloud synchronization for distributed teams
DataForge is being built in the open.
Issues, pull requests, suggestions, and feature requests are welcome.
If you discover something broken or have an idea that could improve DataForge, please open an Issue.
Early feedback directly helps shape what gets built next.
DataForge CLI is licensed under the Apache License 2.0.
See the complete license text in LICENSE.
The open-source core remains available under Apache 2.0. Commercial features are provided through the DataForge Open-Core model.
Developed by RGS Labs™
🌐 Official Website · 💻 GitHub · 🐛 Issues
⭐ If DataForge saves you time, consider starring the repository. It genuinely helps.
Made with ❤️ by RGS Labs™