Skip to main content

Getting Started

Deploy your first MCP server in minutes. This guide walks you through the complete deployment process on agnexus.

Prerequisites​

  • An agnexus account (sign up)
  • An MCP server codebase (or use a template from the marketplace)

Step 1: Start Deployment​

  1. Log in to your agnexus dashboard
  2. Click "Deploy New MCP" button
  3. Choose your deployment method:
    • Marketplace: Browse and deploy pre-built templates
    • GitHub: Deploy from a GitHub repository
    • File Upload: Upload a ZIP file with your code

Step 2: Choose Deployment Method​

Option A: Deploy from Marketplace​

  1. Click "Browse Marketplace"
  2. Browse available MCP templates
  3. Click on a template to view details
  4. Click "Deploy" to use this template
  5. Continue to Step 3

Option B: Deploy from GitHub​

  1. Click "Deploy via GitHub"
  2. Choose repository type:
    • Public Repository: Enter the GitHub repository URL directly
    • Private Repository: Connect your GitHub account first
      • Click "Connect GitHub Account"
      • Authorize agnexus to access your repositories
      • Select your repository from the list
  3. Select the branch you want to deploy (usually main or master)
  4. Continue to Step 3

Option C: Upload ZIP File​

  1. Click "Deploy via File Upload"
  2. Prepare your MCP server code:
    • Zip your project directory (include all files)
    • Make sure it includes requirements.txt (Python) or package.json (Node.js)
    • Include your main entry point file
  3. Click "Choose File" and select your ZIP file
  4. Wait for the file to upload and validate
  5. Continue to Step 3

Step 3: Configure Your MCP Server​

After choosing your deployment method, you'll be asked to configure your server:

1. Server Name​

Enter a name for your MCP server:

  • Use lowercase letters, numbers, and hyphens
  • Must be unique across your account
  • Example: my-notion-mcp or weather-api-server

2. Description (Optional)​

Add a description of what your MCP server does:

  • Helps you identify servers later
  • Visible in your dashboard
  • Example: "Connects AI assistants to my Notion workspace"

3. Environment Variables​

Configure any environment variables your server needs:

  • Click "Add Environment Variable"
  • Enter the variable name (e.g., NOTION_API_KEY)
  • Enter the value (or mark as secret)
  • Add multiple variables as needed

Common environment variables:

  • API keys for external services
  • Database connection strings
  • Configuration flags
  • Service URLs

4. Region​

Select the deployment region:

  • Choose the region closest to your users
  • Affects latency and data residency
  • Can be changed later

5. Dockerfile​

Choose how to handle the Dockerfile:

  • Use Generated Dockerfile: Let agnexus generate a Dockerfile automatically (experimental)
  • Use Custom Dockerfile: Provide your own Dockerfile (recommended for production)
    • If uploading ZIP: Include Dockerfile in your ZIP
    • If using GitHub: Ensure Dockerfile is in your repository

Note: While AI-generated Dockerfiles are available, we strongly recommend using a manual Dockerfile for production deployments as they are more reliable and bulletproof.

Step 4: Deploy​

  1. Review all your configuration settings
  2. Click "Deploy" button
  3. You'll be redirected to the MCP dashboard, specifically to your deployment page

Step 5: Monitor Deployment​

On the deployment page, you can watch your server deploy in real-time:

High-Level Information​

The dashboard shows:

  • Deployment Status: Building, deploying, deployed, or failed
  • Progress Indicators: Visual progress of each deployment stage
  • Current Stage: What's happening right now (e.g., "Building Docker image")
  • Estimated Time: How long the deployment might take

Detailed Information​

For detailed logs and technical information:

  • Click "View Terminal" or "View Logs"
  • See low-level build and deployment logs
  • Debug any issues that occur during deployment

Deployment Stages​

Your deployment goes through these stages:

  1. Validating: Checking your code and configuration
  2. Building: Creating the Docker image
  3. Pushing: Uploading to the container registry
  4. Deploying: Starting your server in production
  5. Health Check: Verifying your server is running

Step 6: Deployment Complete​

Once deployment finishes, you'll see one of two outcomes:

Success ✅​

Your server is now live! You'll see:

  • Deployment Status: "Deployed"
  • Server URL: Your subdomain (e.g., https://my-server.agnexus.ai)
  • Access Information: How to connect to your server

Failure ❌​

If deployment fails:

  • Deployment Status: "Failed"
  • Error Message: What went wrong
  • Logs: Detailed error information in the terminal view
  • Actions: Options to retry or fix issues

Step 7: Access Your Server​

Using the Subdomain​

Your deployed MCP server is accessible at its subdomain:

https://your-server-name.agnexus.ai

You can:

  • Access it directly in a browser (if your server has a web interface)
  • Connect it to MCP clients (like Claude Desktop)
  • Use it in your applications

Adding Access Keys (Optional)​

For private servers, you can add access keys:

  1. Go to your server dashboard
  2. Click "Settings" → "Access Keys"
  3. Click "Generate New Key"
  4. Copy and save your access key
  5. Use it in requests: Authorization: Bearer agn_your_key_here

Next Steps​

Need Help?​