DocuForge

⚡️ Instantly generate clean, intelligent documentation and diagrams for your infrastructure-as-code with the power of LLMs.

---

What is DocuForge?

DocuForge is an open-source CLI tool for automated infrastructure documentation and interactive Q&A. It uses LLMs (local or cloud) to generate Markdown documentation and explain Terraform, Kubernetes YAML, and other config files. You can even ask it questions about your infra, directly from your terminal.

---

✨ Features

• 📄 Automatic Markdown docs for Kubernetes, Terraform, Docker, and more
• 🤖 Interactive chat mode to query your infra like an expert
• 🔐 Local LLM support via Ollama for privacy/offline use
• ☁️ OpenAI integration for enhanced language output (optional)
• ⚙️ Smart regeneration — skip unchanged files
• 🧩 Extensible for other IaC formats and models

---

🚀 Quick Start

1. Clone the Repo

git clone https://github.com/kishanrao92/docuforge.git
cd docuforge

2. Install Requirements

• Go 1.20+
• (Optional) Ollama for local LLMs

3. Run DocuForge

go run main.go --local --model mistral

• Use --local to use a local LLM via Ollama
• Use --model to specify model (e.g., mistral, tinyllama, phi3:mini)

4. Ask Questions

Welcome to DocuForge. Ask me questions about your infrastructure in this repo!
You: What does the deployment.yml file do?
Agent: It defines a Kubernetes Deployment named "dummy" with 3 replicas of the "dummy-container" running image "dummy:0.1" on port 8080.

---

🔍 Why DocuForge?

Infrastructure-as-code grows fast. Docs don’t.
DocuForge helps you:

• Save hours of manual documentation time
• Accelerate onboarding and audits
• Understand infra better through Q&A
• Avoid vendor lock-in with local LLM support
• Stay compliant with up-to-date infra docs

---

📦 Example

Given this file:

apiVersion: apps/v1
kind: Deployment
metadata:
  name: dummy
spec:
  replicas: 3
  template:
    spec:
      containers:
      - name: dummy-container
        image: dummy:0.1
        ports:
        - containerPort: 8080

DocuForge generates:

deployment.yml:
This file defines a Kubernetes Deployment named "dummy" with 3 replicas, running the "dummy-container" (image: dummy:0.1) on port 8080.

---

🖼 Demo

---

⚙️ Configuration

• Place your config files (YAML, HCL, TF, etc.) in a folder like test-configs/
• Output saved to outputs/generated_doc.md
• Set model/provider with CLI flags

---

📚 Supported Providers

Type	Provider	Notes
Local	Ollama	Mistral, TinyLlama, Phi-3, etc.
Cloud	OpenAI	GPT-3.5/4, set OPENAI_API_KEY env

---

🛠 Roadmap

• Markdown generation
• Interactive Q&A mode
• Ollama + OpenAI support
• MermaidJS diagram generation
• GitHub Action integration
• CDK & Pulumi support

---

💡 Ideal For

• DevOps and SRE teams
• Infra/platform engineers
• Startups needing audit-ready docs
• New hires onboarding onto your codebase

---

👏 Contributing

Contributions welcome!
Open issues for features, file type support, or bugs. Star the repo to support the project.

---

📄 License

MIT License

---

🙏 Acknowledgements

• Ollama
• OpenAI
• briandowns/spinner