Custom Servers (MCP)
Custom Servers let you extend Dynaris with your own business logic. Using the Model Context Protocol (MCP), you can deploy tools that your AI agents can call during their workflows.
What is MCP?
The Model Context Protocol is an open standard that allows AI agents to interact with external tool servers. Each MCP server exposes a set of tools -- functions with defined inputs and outputs -- that agents can discover and invoke.
How it works
- You provide a GitHub repository containing your custom tool server code
- Dynaris deploys it as a managed server on dedicated infrastructure
- Agents discover your tools automatically and can use them in conversations
- You manage everything from the Servers page in your dashboard
Creating a custom server
From the Servers page:
- Click Deploy New Server
- Enter the GitHub repository URL containing your MCP server code
- Configure deployment settings (server size, environment variables, etc.)
- Click Deploy and wait for the server to start
Dynaris handles all the infrastructure -- containerization, networking, health checks, and scaling.
Writing an MCP server
Your server exposes tools as functions. Each tool has:
- A name that agents use to invoke it
- A description that helps agents understand when to use it
- Input parameters with types and descriptions
- An output returned to the agent
For example, you might create tools for:
- Querying your internal database
- Calling proprietary APIs
- Running custom business logic
- Generating formatted reports
- Interfacing with legacy systems
Use cases
- Internal APIs -- Give your agents access to your company's proprietary data and services
- Custom workflows -- Encode complex business logic that agents can call as a single tool
- Legacy system integration -- Bridge between Dynaris and older systems that do not have standard API integrations
- Data transformations -- Custom processing that goes beyond what built-in agents can do
- Industry-specific tools -- Build domain-specific capabilities for your particular business
Managing servers
From the Servers page you can:
- View all deployed servers and their status
- Check server health and uptime
- View available tools on each server
- Restart or redeploy servers
- Delete servers you no longer need
Pricing
Custom servers run on dedicated infrastructure. Pricing is based on the server configuration (CPU, memory) and uptime. See Pricing for details.