# MCP Server

Vilna documentation provides a [Model Context Protocol](https://modelcontextprotocol.io/) (MCP) server that gives AI coding assistants direct access to the API reference, endpoint details, and documentation search.

Instead of switching between your IDE and browser to look up endpoint paths or parameter names, your AI assistant can query the live API specification and generate correct integration code on its own.

## Connect your AI tool

The MCP server is available at:


```
https://docs.vilna.io/mcp
```

### Claude Code


```bash
claude mcp add --transport http vilna-docs https://docs.vilna.io/mcp
```

After adding, type `/mcp` in the Claude Code CLI to verify the connection. Claude Code will list the available tools with descriptions and parameters.

### Cursor

1. Open the command palette (`Cmd+Shift+P` on macOS, `Ctrl+Shift+P` on Windows/Linux).
2. Type "Open MCP settings" and select "Add custom MCP".
3. Add to `mcp.json`:



```json
{
  "mcpServers": {
    "vilna-docs": {
      "url": "https://docs.vilna.io/mcp"
    }
  }
}
```

1. Save and confirm the connection in MCP settings.


### VS Code

1. Open the command palette (`Cmd+Shift+P` on macOS, `Ctrl+Shift+P` on Windows/Linux).
2. Type "MCP: Add Server" and select "HTTP".
3. Enter the URL: `https://docs.vilna.io/mcp`
4. Enter a name: `vilna-docs`


Open Chat with AI in Agent mode and select the Tools icon to confirm the connection.

### Claude Desktop

Add to your `claude_desktop_config.json`:


```json
{
  "mcpServers": {
    "vilna-docs": {
      "type": "url",
      "url": "https://docs.vilna.io/mcp"
    }
  }
}
```

### Other tools

Any MCP client that supports the Streamable HTTP transport can connect to `https://docs.vilna.io/mcp`.

## Available tools

Once connected, your AI assistant has access to these tools:

### API discovery

| Tool | Parameters | Description |
|  --- | --- | --- |
| `list-apis` | `filter?` | Lists available APIs (Platform API, Management API) with descriptions |
| `get-endpoints` | `name` | Returns all endpoints for a specific API with descriptions |
| `get-endpoint-info` | `name`, `path`, `method` | Full details for an endpoint - parameters, request body, responses, examples |
| `get-security-schemes` | `name` | Authentication requirements for an API |
| `get-full-api-description` | `name` | Complete OpenAPI specification |


### Documentation search

| Tool | Parameters | Description |
|  --- | --- | --- |
| `search` | `query` | Searches across all documentation and returns relevant content |


## What this looks like in practice

With the MCP server connected, you can ask your AI assistant:

- "How do I register a blockchain address for monitoring with the Vilna API?"
- "What parameters does the create invoice endpoint accept?"
- "Show me the webhook payload format for deposit events"
- "What authentication headers do I need for the Platform API?"
- "Generate a TypeScript function that queries balances for an Ethereum address"


The assistant queries the Vilna MCP server, gets the current API specification, and gives you an accurate answer with correct endpoint paths, parameters, and response schemas - without you leaving your editor.

## Why use the MCP server

**Faster integration.** Your AI assistant knows the exact API surface - every endpoint, every parameter, every response schema. No need to read through documentation pages manually.

**Always current.** The MCP server reads from the live API specification. When endpoints change, your assistant gets the updated information immediately.

**Fewer integration errors.** Instead of guessing parameter names or response formats, the assistant works with the actual OpenAPI specification and generates correct code on the first try.

**Works with your existing tools.** Claude Code, Cursor, VS Code, Claude Desktop - setup takes one command or a small config change.

## Next steps

Quickstart
Make your first API call in 10 minutes

TypeScript SDK
Install the official typed client for the Vilna API

Platform API
Full endpoint reference the MCP server queries