Skip to main content
The tools/call method executes a specific tool with the given arguments. The server validates the arguments against the tool’s input schema, calls the configured API endpoint, and returns the result.

​
MCP Endpoint

POST https://yoursite.com/mcp/{slug}/{server_id}
Content-Type: application/json

​
Request

curl -X POST \
  -H "Content-Type: application/json" \
  -H "Mcp-Session-Id: your_session_id" \
  -d '{
    "jsonrpc": "2.0",
    "id": 4,
    "method": "tools/call",
    "params": {
      "name": "get_weather",
      "arguments": {
        "city": "London"
      }
    }
  }' \
  "https://yoursite.com/mcp/my-weather-tools/a1b2c3d4e5f6a1b2"

​
Response

{
  "jsonrpc": "2.0",
  "id": 4,
  "result": {
    "content": [
      {
        "type": "text",
        "text": "{\"current_condition\":[{\"temp_C\":\"12\",\"temp_F\":\"54\",\"weatherDesc\":[{\"value\":\"Partly cloudy\"}],\"humidity\":\"72\",\"windspeedKmph\":\"15\"}]}"
      }
    ]
  }
}

​
Request Parameters

​
params.name
string
required
The name of the tool to call. Must match a tool defined in the server.
​
params.arguments
object
Key-value pairs matching the tool’s inputSchema. Required parameters must be provided.

​
Response Fields

​
result.content
array
Array of content blocks. Each block has:
  • type — Always "text" in the current implementation
  • text — The response body from the upstream API (JSON string)
​
result.isError
boolean
Present and true when the tool call failed. The content array still contains the error message.

​
Tool Execution Pipeline

When tools/call is received:
  1. Validate — Check the tool exists and is active
  2. Resolve parameters — Map arguments to the HTTP request (path, query, body, headers)
  3. Inject auth — Add API credentials to the request
  4. Call API — Execute the HTTP request to the upstream endpoint
  5. Transform — Convert the API response into MCP content blocks
  6. Log — Record the call in the analytics database
  7. Return — Send the JSON-RPC response