Agents & Tool Calling

Build AI agents that can use tools to accomplish complex tasks.

What You'll Learn

• Overview - What agents are and how they work
• Creating Agents - Configure agent behavior and tools
• Tools - Available built-in and custom tools
• Chatting - Interact with your agents
• Managing - Update and maintain agents
• Tool Calling - Direct tool usage without agents
• Examples - Integration examples

Overview

Agents are AI assistants that can use tools to complete tasks.

How Agents Work

Agents follow a four-step process:

1. Plan - LLM analyzes the task and selects appropriate tools
2. Execute - Run selected tools with provided parameters
3. Observe - Review tool results
4. Synthesize - Combine results into final answer

Agents vs. Raw LLM

Feature	Raw LLM	Agents
Tool usage	No	Yes - Can call APIs, search, execute code
Multi-step tasks	Limited	Yes - Can iterate until task complete
Real-time data	No	Yes - Can fetch current information
Complex workflows	Manual	Yes - Automated tool chains

Use Cases

Use Case	Recommended Tools
Research	RAG search + web search
Data analysis	Code execution
Customer support	RAG + CRM tools
Automation	Custom MCP tools

Quick Example

Create an agent that uses RAG and web search:

import requests

# Create agent
response = requests.post(
    "http://localhost/api/v1/agent/configs",
    headers={"Authorization": "Bearer YOUR_API_KEY"},
    json={
        "name": "research_agent",
        "system_prompt": "You are a research assistant. Use web search for current info and RAG for internal docs.",
        "model": "gpt-4",
        "builtin_tools": ["rag_search", "web_search"],
        "tool_choice": "auto",
        "max_iterations": 5
    }
)

agent_id = response.json()["id"]

# Chat with agent
chat_response = requests.post(
    "http://localhost/api/v1/agent/chat",
    headers={"Authorization": "Bearer YOUR_API_KEY"},
    json={
        "agent_id": agent_id,
        "message": "What are the latest developments in AI?"
    }
)

print(f"Response: {chat_response.json()['response']}")
print(f"Tools used: {chat_response.json()['tool_calls']}")

Next Steps

• Create Your First Agent - Configure an agent
• Available Tools - Learn about built-in and custom tools
• Examples - See how to use agents in applications