EITR — Multi-Provider Terraform Config Generator

---

Introduction

• E — Environment: spanning cloud (AWS, Azure, GCP), local, and virtualised setups
• I — Infrastructure: focused on creating and managing critical infrastructure components
• T — Terraform: the de facto tool for infrastructure as code
• R — Renderer: dynamically generates Terraform configurations

EITR is a web application that simplifies Terraform configuration generation. Select providers and resources from the sidebar, edit the generated HCL templates inline with full syntax highlighting, and download the result as .tf or .json.

In Norse mythology, eitr is a primordial substance of raw creation — the perfect name for a tool that shapes infrastructure.

Live demo: https://eitr.gecko.org

---

Features

• 33 official Terraform providers — AWS, Azure, GCP, Kubernetes, Vault, Consul, and more; all schemas generated directly from the Terraform registry
• Monaco editor — full HCL/Terraform syntax highlighting, bracket matching, and line numbers for inline template editing
• Template persistence — edited templates are saved server-side (templates/custom/) and survive restarts; a "Reset to default" button restores the generated original
• Dark / light theme — toggle persisted to localStorage, defaults to dark
• Resource search — live cross-provider search with 250 ms debounce; click a result to select it directly
• HCL and JSON export — download the assembled configuration in either format; JSON export converts via python-hcl2
• No server-side session state — resource selection is stored in browser localStorage; only custom template edits are written to disk
• Docker Compose — single-command deployment

---

Prerequisites

Docker deployment: Docker and Docker Compose only — schema generation runs inside the build.

Local development: Python 3.12+, Terraform, and jq (for schema generation).

---

Quick start — Docker Compose

Schema generation runs automatically during the Docker build. No local Terraform required.

git clone https://github.com/daytonjones/EITR.git
cd EITR
docker compose up --build

Open http://localhost:8085 in your browser.

First build is slow — Terraform downloads provider plugins (~500 MB). Subsequent builds use Docker's layer cache and are fast; the provider download only re-runs when config/providers.json changes.

To update a running deployment:

git pull origin main
docker compose up --build -d

---

Local development

git clone https://github.com/daytonjones/EITR.git
cd EITR

python3 -m venv venv
source venv/bin/activate        # Windows: venv\Scripts\activate
pip install -r requirements.txt

# Generate provider schemas and Jinja2 templates (required before first run)
python utilities/generate_tf_provider_templates.py

uvicorn main:app --reload

Open http://localhost:8000 in your browser.

---

Schema generation

utilities/generate_tf_provider_templates.py must be run:

• Before starting the app for the first time after a fresh clone
• After modifying config/providers.json (adding/removing/updating providers)

The script:

1. Writes a temporary main.tf from config/providers.json
2. Runs terraform init to download provider plugins (slow on first run)
3. Runs terraform providers schema -json | jq . and saves the output to config/provider_schemas.json
4. Generates Jinja2 template skeletons for every resource, data source, ephemeral resource, function, and provider-config block into templates/terraform/{provider}/

config/provider_schemas.json is gitignored (it is large); the generated templates in templates/terraform/ are committed.

---

Adding or updating providers

Edit config/providers.json. Each entry requires:

{ "name": "aws", "description": "AWS (Amazon Web Services)", "version": "5.78.0" }

For providers not under the hashicorp/ namespace, add a source field:

{ "name": "vsphere", "description": "VMware vSphere", "version": "2.10.0", "source": "vmware/vsphere" }

Then re-run the generator.

---

Usage

1. Open the app in your browser.
2. Use the sidebar to expand a provider and check the resources you need.
3. Use the search box to find a specific resource across all providers.
4. Click Edit on any resource card to edit its template in the Monaco editor.
5. Click Save to persist your edits to the server; a "customized" badge appears on the card.
6. Click Reset to default to restore the generated template.
7. Click HCL or JSON in the action bar to download the assembled configuration.
8. Click Clear all to deselect all resources (custom template edits are preserved).

---

Project structure

EITR/
├── config/
│   └── providers.json           # Provider list (source of truth)
│   # provider_schemas.json      # Generated — gitignored
├── templates/
│   ├── index.html               # Single-page frontend
│   ├── terraform/               # Generated .tf.j2 skeletons (committed)
│   │   ├── aws/
│   │   ├── azurerm/
│   │   └── …
│   └── custom/                  # User-edited overrides (committed, volume-mounted)
├── static/
│   ├── eitr_background.jpeg
│   ├── eitr.ico
│   └── eitr.png
├── utilities/
│   └── generate_tf_provider_templates.py
├── main.py
├── requirements.txt
├── Dockerfile
├── docker-compose.yml
└── CHANGELOG.md

---

License

EITR is licensed under the MIT License. See LICENSE for details.