AI-driven Dockerfile, DevOps automation, and CI/CD orchestrator for any Python project.
Orchify (formerly Orchestron) is a command-line tool that uses AI agents (via Agno and the Model Context Protocol) to:
- Scan a Python project and analyze its structure, dependencies, entrypoints, and infrastructure needs.
- Generate production-grade, multi-stage Dockerfiles, .dockerignore, Docker Compose, CI/CD workflows, and other DevOps configurations—tailored to your project automatically.
Whether you’re containerizing microservices, setting up cloud deployments, or building full CI/CD pipelines, Orchify crafts best-practice infrastructure in seconds.
- Framework Agnostic: Works with plain Python scripts, Flask, FastAPI, Django, AWS Lambdas, and more.
- AI-Powered: Leverages Anthropic Claude or OpenAI GPT models to write idiomatic DevOps artifacts.
- Extensible: Generates Dockerfiles,
docker-compose.yml, GitHub Actions workflows, and can be extended for Terraform, Kubernetes, etc. - Token-Aware: Dynamically selects models based on codebase size to avoid context limits.
- Typed Manifest: Produces a Pydantic-validated
orchify.jsonas a single source of truth for your DevOps pipeline. - Logging & Debugging: Built-in logging to trace scans and generations.
pip install orchifyOr from source:
git clone https://github.com/sattyamjjain/Orchify.git
cd orchify
pip install -e .Orchify CLI has three main commands:
A quick check:
orchify hello
# 👋 Welcome to Orchify!Analyze and build orchify.json manifest:
orchify scan --dir /path/to/your/projectOptions:
-d, --dirpath: Project directory (default: current).-p, --prompttext: Custom AI prompt.
Generate DevOps artifacts from the manifest:
orchify gen --model-id claude-3-7-sonnet-latest --output ./DockerfileOptions:
-m, --model-idmodel: AI model.-o, --outputpath: Dockerfile path (default:Dockerfile).
Authenticate AI providers via environment variables:
| Variable | Description |
|---|---|
ANTHROPIC_API_KEY |
Anthropic Claude API token. |
OPENAI_API_KEY |
OpenAI API key. |
Orchify emits a Pydantic-validated manifest that includes:
- Project metadata (name, version, repo)
- Dependencies & install commands
- Source & entrypoint files
- Docker Compose service definitions
- CI/CD workflow settings
- Security scanning configs
- Notification hooks
Use this manifest to drive CI/CD or further automation.
pip install -e .
pytest tests/
flake8 orchify tests
black orchify testsContributions welcome! Please fork, branch, and open a PR.
MIT License — see LICENSE for details.