diagram-to-iac 1.0.3

Convert architecture diagrams into IaC modules

diagram-to-iac

"One container, many minds—zero manual toil."

An automated DevOps-in-a-Box system that intelligently handles complete Repo-to-Deployment (R2D) workflows. The project combines AI-powered infrastructure analysis with GitHub automation for self-healing deployments.

🚀 DevOps-in-a-Box: R2D Action

The R2D (Repo-to-Deployment) Action is a self-healing, Terraform-first DevOps automation solution that lives inside a single GitHub Action. When you supply any repository URL, our SupervisorAgent marshals specialized agents to handle the complete deployment workflow.

Quick Start with GitHub Actions

name: Deploy Infrastructure
on:
  push:
    branches: [main]

jobs:
  deploy:
    runs-on: ubuntu-latest
    steps:
      - name: Deploy with R2D
        uses: amartyamandal/diagram-to-iac/.github/actions/r2d@v1
        with:
          repo: ${{ github.server_url }}/${{ github.repository }}
        env:
          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
          TF_CLOUD_TOKEN: ${{ secrets.TF_CLOUD_TOKEN }}

Key Features

• 🤖 Self-Healing: Automatically creates GitHub Issues for errors and suggests fixes
• 🔒 Security-First: Non-root container execution with workspace isolation
• 🌍 Multi-Cloud: Supports Terraform, PowerShell, Bash, and Ansible deployments
• 📊 Observable: Rich logging, step summaries, and artifact collection
• 🔄 Resumable: Thread-based conversation tracking for workflow continuation
• 🧪 Testable: Comprehensive dry-run mode for safe testing

The Cast: Specialized Agents

Agent	Capability	Never Does
SupervisorAgent	Orchestrates workflow, manages checkpoints	Edit code directly
GitAgent	Clone, branch, PR creation, assign @github-copilot	Guess network credentials
ShellAgent	Safe command execution, stack detection	Execute non-allowlisted binaries
TerraformAgent	Init/plan/apply, error classification	Apply with critical security issues
PolicyAgent	tfsec + OPA security gates	Ignore critical findings

📦 Installation

Create a virtual environment with Python 3.11+ and install the project in editable mode. Development dependencies (linting, testing, etc.) are provided through the dev extra:

pip install -e .[dev]

🖥️ Running the CLI

The project exposes several entry points via pyproject.toml:

Main R2D CLI

# Run complete R2D workflow
diagram-to-iac https://github.com/user/repo
r2d-agent https://github.com/user/repo --dry-run

# Get help
diagram-to-iac --help

Individual Agent CLIs

# SupervisorAgent (orchestration)
supervisor-agent --repo-url https://github.com/user/repo

# GitAgent (repository operations)
git-agent --repo-url https://github.com/user/repo

# TerraformAgent (infrastructure deployment)
terraform-agent --query "terraform plan"

Interactive Mode

Running without arguments enters interactive mode:

$ supervisor-agent --dry-run
╔══════════════════════════════════════════════════════════════════════════════╗
║  🤖 DevOps-in-a-Box: SupervisorAgent                                        ║
║  "One container, many minds—zero manual toil."                              ║
╚══════════════════════════════════════════════════════════════════════════════╝

Repository URL: https://github.com/octocat/Hello-World.git
🚀 R2D SupervisorAgent - Workflow Orchestration
📅 Default branch name: r2d-<timestamp>
📝 Press Enter to use default, or type a custom branch name:

The agent will continue with the complete workflow (clone → detect → deploy → issue creation). The --dry-run flag simulates actions without making changes.

Running Tests

All tests use pytest and are located under the tests directory. After installing the development dependencies, run:

pytest

📊 Logs and Observability

DevOps-in-a-Box provides comprehensive observability for all R2D workflows:

Structured Logging

Each run creates a JSONL log file under the logs/ directory (e.g. logs/run-<timestamp>.jsonl). Every significant event is logged as structured JSON for easy parsing and analysis:

# Follow live logs
tail -f logs/run-*.jsonl

# Parse with jq for specific events
cat logs/run-*.jsonl | jq '. | select(.event_type == "terraform_apply")'

Step Summary Dashboard

After each workflow, a Markdown dashboard is generated at step-summary.md with:

• 📈 Terraform resource changes and cost estimates
• 🔒 Security findings from tfsec/OPA scans
• 🏗️ Infrastructure modules and dependencies
• ⚡ Performance metrics and execution times

GitHub Integration

• Issues: Automatically created for errors with detailed context
• Pull Requests: Auto-drafted fixes assigned to @github-copilot
• Workflow Summaries: Rich GitHub Actions step summaries
• Artifacts: Logs, plans, and reports uploaded for download

CI/CD Artifacts

In GitHub Actions, the following artifacts are automatically collected:

• logs/ - Structured JSONL logs
• *.tfplan - Terraform plan files
• step-summary.md - Workflow dashboard
• r2d-artifacts/ - Complete workflow artifacts

🔧 GitHub Actions Usage

Add the R2D action to your workflow for automated infrastructure deployment:

Basic Setup

name: Deploy Infrastructure
on:
  push:
    branches: [main]

jobs:
  deploy:
    runs-on: ubuntu-latest
    steps:
      - name: Deploy with R2D
        uses: amartyamandal/diagram-to-iac/.github/actions/r2d@v1
        with:
          repo: ${{ github.server_url }}/${{ github.repository }}
        env:
          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
          TF_CLOUD_TOKEN: ${{ secrets.TF_CLOUD_TOKEN }}

Advanced Configuration

- name: Deploy with Custom Settings
  uses: amartyamandal/diagram-to-iac/.github/actions/r2d@v1
  with:
    repo: 'https://github.com/my-org/infrastructure'
    package_version: 'v2.1.0'
    dry_run: 'false'
    branch_name: 'deploy-prod'
    thread_id: 'prod-deployment-001'
  env:
    GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
    TF_CLOUD_TOKEN: ${{ secrets.TF_CLOUD_TOKEN }}
    OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
    ANTHROPIC_API_KEY: ${{ secrets.ANTHROPIC_API_KEY }}

Required Secrets

Configure these in your repository settings:

Secret	Description	Required
GITHUB_TOKEN	GitHub API access (auto-provided)	✅ Yes
TF_CLOUD_TOKEN	Terraform Cloud API token	✅ Yes
OPENAI_API_KEY	OpenAI API key	❌ Optional
ANTHROPIC_API_KEY	Claude API key	❌ Optional
GOOGLE_API_KEY	Gemini API key	❌ Optional

Example Workflows

See .github/actions/r2d/examples/ for complete workflow examples:

• Basic Deployment: Simple push-to-deploy workflow
• Multi-Environment: Deploy to dev/staging/prod with approvals
• PR Validation: Validate infrastructure changes in pull requests
• Drift Detection: Scheduled infrastructure drift monitoring
• External Repository: Deploy from external repositories

After each workflow run, a Markdown dashboard is generated at step-summary.md showing a high level overview of Terraform modules, resource changes and tfsec findings. The dashboard is derived from the JSONL logs and can be viewed directly in the repository or uploaded as a build artifact.

This repository provides a container action that runs the SupervisorAgent on the current repository. Add the action to a workflow as shown below:

jobs:
  supervisor:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - name: Run Supervisor Agent
        uses: ./.github/actions/supervisor
        env:
          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}

The action reads GITHUB_REPOSITORY and GITHUB_TOKEN automatically to clone the repository and execute the agent.