MCP

Model Context Protocol operations for tool discovery and execution.

List available tools

get

Retrieves a list of available tools (capabilities) for the specific AI app deployment.

Authorizations

ApiKeyAuth

nlx-api-keystringRequired

Path parameters

deploymentKeystringRequired

The unique key identifying the AI app deployment.

channelKeystringRequired

The unique key identifying the specific channel. Must include the language code suffix (e.g., -en-US, -es-MX).

Example: xxxxxxxx-en-US

Responses

200

A list of tools.

application/json

401

Unauthorized - Invalid API Key

get

/c/mcp/{deploymentKey}/{channelKey}/tools

HTTPcURLJavaScriptPython

GET /c/mcp/{deploymentKey}/{channelKey}/tools HTTP/1.1
Host: apps.nlx.ai
nlx-api-key: YOUR_API_KEY
Accept: */*

{
  "tools": [
    {
      "name": "text",
      "description": "text",
      "inputSchema": {
        "ANY_ADDITIONAL_PROPERTY": "anything"
      }
    }
  ]
}

Execute a tool

post

Executes a specific tool by name. Supports both standard JSON responses and Server-Sent Events (SSE) streaming if requested.

Authorizations

ApiKeyAuth

nlx-api-keystringRequired

Path parameters

deploymentKeystringRequired

The unique key identifying the AI app deployment.

channelKeystringRequired

The unique key identifying the specific channel. Must include the language code suffix (e.g., -en-US, -es-MX).

Example: xxxxxxxx-en-US

toolNamestringRequired

The name of the tool to execute.

Body

application/json

streambooleanOptional

Set to true to receive an SSE stream of the tool's output.

Default: false

Responses

200

Tool execution result.

application/jsontext/event-stream

400

Bad Request

401

Unauthorized

post

/c/mcp/{deploymentKey}/{channelKey}/tools/{toolName}

HTTPcURLJavaScriptPython

POST /c/mcp/{deploymentKey}/{channelKey}/tools/{toolName} HTTP/1.1
Host: apps.nlx.ai
nlx-api-key: YOUR_API_KEY
Content-Type: application/json
Accept: */*
Content-Length: 67

{
  "arguments": {
    "ANY_ADDITIONAL_PROPERTY": "anything"
  },
  "stream": false
}

{
  "content": [
    {
      "type": "text",
      "text": "text",
      "data": "text",
      "mimeType": "text"
    }
  ],
  "isError": false
}