# Publishing and Using infranodus-mcp-server with npx
## Publishing to npm
1. **Ensure you're logged in to npm:**
```bash
npm login
```
2. **Build the project:**
```bash
npm run build:inspect
```
3. **Publish to npm:**
```bash
npm publish
```
Or for a test run first:
```bash
npm publish --dry-run
```
## Using with npx
Once published, users can run your MCP server using npx in their Claude Desktop or other MCP client configurations:
### Claude Desktop Configuration
Users can add this to their Claude Desktop config file (`claude_desktop_config.json`):
```json
{
"mcpServers": {
"infranodus": {
"command": "npx",
"args": [
"-y",
"infranodus-mcp-server"
],
"env": {
"INFRANODUS_API_KEY": "your-api-key-here",
"INFRANODUS_API_BASE": "https://infranodus.com/api/v1"
}
}
}
}
```
### Direct Testing with npx
After publishing, test your package:
```bash
# Run directly (will exit immediately as it expects MCP protocol)
npx -y infranodus-mcp-server
# Set environment variables if needed
INFRANODUS_API_KEY=your-key npx -y infranodus-mcp-server
```
## Version Updates
When you make changes:
1. Update the version in `package.json`:
```bash
npm version patch # for bug fixes
npm version minor # for new features
npm version major # for breaking changes
```
2. Build and publish:
```bash
npm run build:inspect
npm publish
```
## Local Testing Before Publishing
To test the npx behavior locally before publishing:
1. Build the project first:
```bash
npm run build:inspect
```
2. Link your package globally:
```bash
npm link
```
3. Test the command with a simple initialize message:
```bash
echo '{"jsonrpc":"2.0","id":1,"method":"initialize","params":{"protocolVersion":"2024-11-05","capabilities":{},"clientInfo":{"name":"test","version":"1.0.0"}}}' | infranodus-mcp-server
```
You should see a JSON response with server capabilities.
4. Unlink when done:
```bash
npm unlink -g infranodus-mcp-server
```
## Troubleshooting
If the server disconnects unexpectedly:
1. Check that environment variables are set correctly
2. Ensure the build is up to date (`npm run build:inspect`)
3. Check stderr output by adding debugging statements with `console.error()`
## Package Structure
The published package includes:
- `dist/` - Compiled JavaScript files
- `bin/` - Executable entry point with shebang
- `README.md` - Documentation
- `LICENSE` - MIT License
- `package.json` - Package metadata
Files excluded (via `.npmignore`):
- Source TypeScript files (`src/`)
- Development configuration files
- Environment files (`.env`)
- Development dependencies
---
## Deploying to Fly.io (HTTP Server)
The MCP server can be deployed as an HTTP server with OAuth2 authentication at a public URL (e.g., `mcp.infranodus.com`).
### Prerequisites
1. Install the Fly.io CLI:
```bash
brew install flyctl
# or
curl -L https://fly.io/install.sh | sh
```
2. Login to Fly.io:
```bash
fly auth login
```
### Initial Deployment
1. **Create the app** (first time only):
```bash
fly launch --no-deploy
```
This creates the app based on `fly.toml` configuration.
2. **Set secrets**:
```bash
fly secrets set JWT_SECRET="your-secure-random-secret"
```
Generate a secure secret:
```bash
openssl rand -base64 32
```
3. **Deploy**:
```bash
fly deploy
```
### Subsequent Deployments
After the initial setup, deploy updates with:
```bash
npm run build:inspect
fly deploy
```
### Custom Domain Setup
To use a custom domain like `mcp.infranodus.com`:
1. **Add the domain**:
```bash
fly certs add mcp.infranodus.com
```
2. **Configure DNS** - Add the records shown by Fly.io to your domain's DNS settings (typically a CNAME to `infranodus-mcp-server.fly.dev`).
3. **Verify**:
```bash
fly certs show mcp.infranodus.com
```
### Configuration
The `fly.toml` configures:
- **Region**: Frankfurt (`fra`)
- **Resources**: 1GB RAM, 1 shared CPU
- **Always-on**: Minimum 1 machine running
- **Health checks**: Every 30s on `/health`
Environment variables set in `fly.toml`:
```toml
[env]
CORS_ORIGIN = '*'
INFRANODUS_API_BASE = 'https://infranodus.com/api/v1'
NODE_ENV = 'production'
```
### Monitoring
```bash
# View logs
fly logs
# Check app status
fly status
# Open the app
fly open
# SSH into the container
fly ssh console
```
### Scaling
```bash
# Scale to multiple regions
fly scale count 2 --region fra,ams
# Adjust VM size
fly scale vm shared-cpu-1x --memory 1024
```
### Alternative: Render.com
A `render.yaml` is also provided for deploying to Render.com:
1. Connect your GitHub repo to Render
2. Render will auto-detect the `render.yaml` blueprint
3. Set the `JWT_SECRET` environment variable in the Render dashboard
