Google AI Studio (Gemini)

Google AI Studio supports external function calling. For detailed configuration instructions, visit the official documentation.

To integrate Timbr:

Function Declarations

Define Timbr tools as function declarations in your AI Studio project:

const timbrTools = [
  {
    name: "query_data",
    description: "Execute a natural language query against Timbr semantic layer",
    parameters: {
      type: "object",
      properties: {
        prompt: {
          type: "string",
          description: "Natural language query"
        },
        agent_name: {
          type: "string",
          description: "Target agent name"
        },
        ontology: {
          type: "string",
          description: "Target ontology name"
        },
        max_limit: {
          type: "integer",
          description: "Maximum rows to return"
        }
      },
      required: ["prompt"]
    }
  },
  {
    name: "ask_question",
    description: "Ask a question and get a natural language answer",
    parameters: {
      type: "object",
      properties: {
        prompt: {
          type: "string",
          description: "Natural language question"
        },
        agent_name: {
          type: "string",
          description: "Target agent name"
        },
        ontology: {
          type: "string",
          description: "Target ontology name"
        }
      },
      required: ["prompt"]
    }
  }
];

Function Implementation

Implement the function handler to call Timbr API:

async function handleTimbrFunction(functionCall) {
  const { name, args } = functionCall;
  
  const headers = {
    'Content-Type': 'application/json',
    'x-api-key': process.env.TIMBR_API_KEY
  };
  
  // Set either agent or ontology header
  if (args.agent_name) {
    headers['x-agent'] = args.agent_name;
  } else if (args.ontology) {
    headers['x-ontology'] = args.ontology;
  }
  
  const response = await fetch('https://<your-timbr-server>/timbr/api/mcp/', {
    method: 'POST',
    headers: headers,
    body: JSON.stringify({
      jsonrpc: '2.0',
      id: 1,
      method: 'tools/call',
      params: {
        name: name,
        arguments: args
      }
    })
  });
  
  return await response.json();
}