> For the complete documentation index, see [llms.txt](https://whitepaper.aitech.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://whitepaper.aitech.io/agentforge/introduction/getting-started.md).

# Getting Started

{% hint style="info" %}
This tutorial takes about 10 minutes and covers the essential concepts of building workflows in Agent Forge.
{% endhint %}

This tutorial will guide you through building your first AI workflow in Agent Forge. We'll create a people research agent that can find information about individuals using state-of-the-art LLM-Search tools.

### What We're Building

A people research agent that:

1. Receives a person's name via chat interface
2. Uses an AI agent with advanced search capabilities
3. Searches the web using state-of-the-art LLM-Search tools (Exa)
4. Extracts structured information using a response format
5. Returns comprehensive data about the person

### Step-by-Step Tutorial

{% stepper %}
{% step %}

### Create workflow and add AI agent

Open Agent Forge and click "New Workflow" in the dashboard. Name it "Getting Started".

When you create a new workflow, it automatically includes a **Start block** - this is the entry point that receives input from users. For this example, we'll be triggering the workflow via chat, so we don't need to configure anything on the Start block.

Now drag an **Agent Block** onto the canvas from the blocks panel on the left.

Configure the Agent Block:

* **Model**: Select "OpenAI GPT-4o"
* **System Prompt**: "You are a people research agent. When given a person's name, use your available search tools to find comprehensive information about them including their location, profession, educational background, and other relevant details."
* **User Prompt**: Drag the connection from the Start block's output into this field (this connects `<start.input>` to the user prompt)
  {% endstep %}

{% step %}

### Add tools to the agent

Let's enhance our agent with tools for better capabilities. Click on the Agent block to select it.

In the **Tools** section:

* Click **Add Tool**
* Select **Exa** and **Linkup** from the available tools
* Add your API keys for both tools (this allows the agent to search the web and access additional information)
  {% endstep %}

{% step %}

### Test the basic workflow

Now let's test our workflow. Go to the **Chat panel** on the right side of the screen.

In the chat panel:

* Click the dropdown and select `agent1.content` (this will show us the output of our agent)
* Enter a test message like: "Who is Nikola Tesla?"
* Click "Send" to run the workflow

You should see the agent's response analyzing the person described in your text.
{% endstep %}

{% step %}

### Add structured output

Now let's make our agent return structured data. Click on the Agent block to select it.

In the **Response Format** section:

* Click the **magic wand icon** (✨) next to the schema field
* In the prompt that appears, type: "create a schema named person, that contains location, profession, and education"
* The AI will generate a JSON schema for you automatically
  {% endstep %}

{% step %}

### Test the structured output

Go back to the **Chat panel**.

Since we added a response format, new output options are now available:

* Click the dropdown and select the new structured output option (the schema we just created)
* Enter a new test message like: "Who is Alan Turing?"
* Click "Send" to run the workflow again

You should now see structured JSON output with the person's information organized into location, profession, and education fields.
{% endstep %}
{% endstepper %}

### What You Just Built

Congratulations! You've created your first AI workflow that:

* ✅ Receives text input via chat interface
* ✅ Uses AI to extract information from unstructured text
* ✅ Integrates external tools (Exa) for enhanced capabilities
* ✅ Returns structured JSON data using AI-generated schemas
* ✅ Demonstrates workflow testing and iteration
* ✅ Shows the power of visual workflow building

### Key Concepts You Learned

#### Block Types Used

| Title       | Target                                     |
| ----------- | ------------------------------------------ |
| Start Block | Entry point for user input (auto-included) |
| Agent Block | AI model for text processing and analysis  |

#### Core Workflow Concepts

**Data Flow**: Variables flow between blocks by dragging connections

**Chat Interface**: Test workflows in real-time using the chat panel with different output options

**Tool Integration**: Enhance agent capabilities by adding external tools like Exa

**Variable References**: Access block outputs using `<blockName.output>` syntax

**Structured Output**: Use JSON schemas to get consistent, structured data from AI

**AI-Generated Schemas**: Use the magic wand (✨) to generate schemas with natural language

**Iterative Development**: Test, modify, and re-test workflows easily


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://whitepaper.aitech.io/agentforge/introduction/getting-started.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.