OpenAI Tools Agent

The OpenAI Tools Agent provides a powerful interface for leveraging OpenAI's function calling capabilities with custom tools. It manages tool execution, conversation context, and provides intelligent responses while maintaining error handling and conversation history.

OpenAI Tools Agent workflow and architecture

Configuration Parameters

Required Input Parameters

tools: Array of tool configurations
input: User query or instruction
system_prompt: Base system instructions

Optional Configuration

llm: Language model configuration
- model_name: Name of the OpenAI model
- temperature: Response creativity (0-1)
- max_tokens: Maximum response length
agent_config: Agent behavior settings
- agent_description: Role and capabilities
- handle_parse_errors: Error handling strategy
- verbose: Detailed logging
- max_iterations: Maximum tool call iterations
chat_history: Previous conversation context

Output Format

{
  "agent": {
    "status": "success" | "error",
    "tools_used": [
      {
        "name": string,
        "calls": number,
        "success_rate": number
      }
    ],
    "iterations": number,
    "processing_time": number
  },
  "response": {
    "message": string,
    "tool_outputs": [
      {
        "tool": string,
        "input": object,
        "output": any,
        "timestamp": string
      }
    ],
    "reasoning": string,
    "next_steps": array
  },
  "context": {
    "chat_history": array,
    "system_prompt": string,
    "current_state": object
  },
  "metadata": {
    "model_info": {
      "name": string,
      "temperature": number
    },
    "token_usage": {
      "prompt": number,
      "completion": number,
      "total": number
    }
  }
}

Features

Custom tool integration
Conversation management
Error handling and recovery
Tool execution tracking
Context preservation
Performance monitoring
Token usage optimization
Detailed logging

Note: Tool definitions should be clear and specific. Consider token limits when designing tool inputs and outputs.

Tip: Enable verbose mode during development for better debugging. Use appropriate max_iterations to prevent infinite tool calling loops.

Example Usage

const toolsAgent = new OpenAIToolsAgent({
  tools: [
    {
      name: "calculator",
      description: "Perform mathematical calculations",
      parameters: {
        expression: "string"
      }
    },
    {
      name: "weather",
      description: "Get weather information",
      parameters: {
        location: "string",
        unit: "celsius" | "fahrenheit"
      }
    }
  ],
  llm: {
    model_name: "gpt-4",
    temperature: 0.3
  },
  system_prompt: "You are a helpful assistant with access to various tools...",
  agent_config: {
    agent_description: "Multi-tool processing assistant",
    handle_parse_errors: true,
    verbose: true,
    max_iterations: 5
  }
});

const response = await toolsAgent.process({
  input: "What's the temperature in London and convert it to Fahrenheit?",
  chat_history: previousMessages
});