StrixReady CLI

Instant dev environment from any GitHub URL — run it locally or containerise it.

Give Strix a GitHub repo link — it clones the repo, analyses the stack, and uses AI to either:

• Run it instantly on localhost (AI infers install + dev commands, starts the dev server)
• Generate production-ready Docker config files (Dockerfile, docker-compose.yml, .dockerignore, .env.example)

How It Works

Run (local dev)

GitHub URL → Clone → Analyse (package.json / requirements.txt / README)
     → AI returns install + dev commands → Execute locally → App on localhost

Generate (Docker)

GitHub URL → Clone → Analyse → AI generates Dockerfile + docker-compose.yml
     → Files written to cloned repo → Ready to docker compose up

Quick Start

# Install
pip install -e .

# Set your Groq API key
echo "GROQ_API_KEY=gsk_your_key_here" > .env

# Run a repo locally
strix scan https://github.com/owner/repo

# Start the GUI backend (React frontend on port 8080 calls this)
strix gui

CLI Commands

Command	Description
strix scan <url>	Clone, detect stack, install deps, run dev server locally
strix scan <url> --os macos	Specify target OS
strix gui	Start the backend API (port 8000) for the React frontend
strix doctor	Health-check running services

API Endpoints (for React frontend)

Method	Endpoint	Description
GET	/health	Health check
POST	/scan	Generate — Clone + analyse + AI generates Docker config files
GET	/scan/stream	Run — Clone + analyse + execute locally with SSE live progress
POST	/scan/analyze	Clone + analyse only (preview commands, no execution)
POST	/run	Execute commands in a given local directory

---

POST /scan — Docker file generation (Generate button)

Request:

{ "url": "https://github.com/owner/repo", "os": "linux" }

Response:

{
  "profile": {
    "name": "repo",
    "languages": ["TypeScript", "JavaScript"],
    "frameworks": ["Vite", "React"],
    "local_path": "/tmp/strix_.../repo"
  },
  "artifacts": {
    "Dockerfile": "FROM node:20-alpine\n...",
    "docker-compose.yml": "services:\n  repo:\n...",
    ".dockerignore": "node_modules\n.git\n...",
    ".env.example": "# API keys\nVITE_API_URL=...",
    "notes": "Single-stage Vite build. Run: docker compose up"
  },
  "written_files": [
    "/tmp/strix_.../repo/Dockerfile",
    "/tmp/strix_.../repo/docker-compose.yml",
    "/tmp/strix_.../repo/.dockerignore",
    "/tmp/strix_.../repo/.env.example"
  ],
  "local_path": "/tmp/strix_.../repo"
}

---

GET /scan/stream — Local dev server with live progress (Run button)

SSE stream — each event is a JSON object:

GET /scan/stream?url=https://github.com/owner/repo&os=linux

Events:

data: {"step": "clone",       "message": "Cloning repository...",           "data": null}
data: {"step": "analyze",     "message": "Analysis complete",               "data": {...}}
data: {"step": "ai",          "message": "AI returned commands",            "data": {...}}
data: {"step": "commands",    "message": "Setup plan ready",                "data": {"install_command": "npm i", "dev_command": "npm run dev", ...}}
data: {"step": "install",     "message": "Installing dependencies: npm i",  "data": null}
data: {"step": "install",     "message": "Dependencies installed.",         "data": null}
data: {"step": "dev",         "message": "Starting dev server: npm run dev","data": null}
data: {"step": "done",        "message": "App running at http://localhost:8081", "data": {"running": true, "port": 8081, "pid": 12345}}
data: {"step": "end",         "message": "Stream complete",                 "data": null}

Steps: clone → analyze → ai → commands → pre_install → install → post_install → dev → done → end

---

POST /scan/analyze — Preview only (no execution)

Request:

{ "url": "https://github.com/owner/repo", "os": "linux" }

Response:

{
  "profile": { "name": "repo", "languages": ["Python"], "frameworks": ["FastAPI"] },
  "commands": {
    "install_command": "pip install -r requirements.txt",
    "dev_command": "uvicorn main:app --reload",
    "port": 8000,
    "env_vars": { "DATABASE_URL": "" },
    "env_notes": "Set DATABASE_URL to your Postgres connection string."
  }
}

---

POST /run — Execute commands in a local directory

Request:

{
  "local_path": "/tmp/strix_.../repo",
  "commands": {
    "install_command": "npm install",
    "dev_command": "npm run dev",
    "port": 5173
  }
}

Response:

{ "result": { "running": true, "port": 5173, "pid": 12345, "error": null } }

---

Project Structure

├── cli/main.py            # Typer CLI (scan, gui, doctor)
├── backend/
│   ├── main.py            # FastAPI app (port 8000)
│   ├── analyzer.py        # Clone repo + detect stack (with on_log callbacks)
│   ├── generator.py       # AI command/Docker generation + local execution
│   ├── commands.py        # Deterministic fallback command inference
│   ├── health.py          # Health checks
│   └── utils.py           # Shared colour constants
├── prompts/
│   ├── generate_artifacts_prompt.txt  # System prompt for local dev commands (Run)
│   └── generate_docker_prompt.txt     # System prompt for Docker file generation (Generate)
├── pyproject.toml
└── .env                   # GROQ_API_KEY

Requirements

• Python 3.10+
• Git
• Node.js / npm (for JS/TS projects)
• Groq API key (free at console.groq.com)

License

MIT .venv/bin/uvicorn backend.main:app --host 0.0.0.0 --port 8000 --reload