# AutoGen

## 1. Install Required Libraries

First, you need to install the necessary packages:

{% hint style="info" %}
Integry requires Python version 3.12 or higher
{% endhint %}

* **Integry** is used to integrate structured tools and functions.
* **AutoGen** integrate tools and automate workflows using large language models.

```python
pip install integry autogen
```

## 2. Initialize Integry & Agent

Import the necessary Libraries

```python
import os
from integry import Integry
from autogen import ConversableAgent, register_function
```

`User-ID` is a unique string identifier for a user in your app or agent. Function Calls and Integrations are associated to a user ID. It will be the email address you used during the signup process on Integry.

For example:

```python
user_id = "joe@example.com"
```

Below code snippet initializes the **Integry** class to interact with the Integry API using the **App-Key** and **App-Secret**.

You can view and copy your `App-Key` and `App-Secret` from the [Workspace Settings](https://app.integry.io/platform/workspace/security/).

<figure><img src="/files/gXJ7oZzZ0UMrM6pfWAlg" alt=""><figcaption><p>Workspace Settings</p></figcaption></figure>

```python
integry = Integry(
    app_key=os.environ.get("INTEGRY_APP_KEY"),
    app_secret=os.environ.get("INTEGRY_APP_SECRET"),
)
```

The code creates two AI agents: one is "Assistant" that helps perform tasks, and another is "User" that listens for a **`TERMINATE`** message to stop.

The assistant uses an API key for GPT-4 You can get the API Key from the [OpenAI Platform](https://platform.openai.com/api-keys), while the user agent has a rule to stop when it receives the **`TERMINATE`** message.

<pre class="language-python" data-overflow="wrap"><code class="lang-python"><strong>llm_config = {"config_list": [{"model": "gpt-4o", "api_key": os.environ.get("OPENAI_API_KEY")}]}
</strong>
assistant = ConversableAgent(
    name="Assistant",
    system_message="You are a helpful integrations assistant. "
    "You can help users perform tasks in various apps. "
    "Return 'TERMINATE' when the task is done.",
    llm_config=llm_config,
)

user_proxy = ConversableAgent(
    name="User",
    llm_config=False,
    is_termination_msg=lambda msg: msg.get("content") is not None
    and "TERMINATE" in msg["content"],
    human_input_mode="NEVER",
)
</code></pre>

## 3. Register an Integry Function as a Tool

Before you can use the functions available in Integry, you need to add the app to Integry. Slack, however, is pre-added to Integry by default, so there’s no need to add it manually.

The code retrieves a Slack post-message function from [Integry](https://app.integry.io/platform/functions). You can copy the function ID from the dropdown.

For example

In this case the function ID is <mark style="color:blue;">slack-post-message</mark>

<figure><img src="/files/fZytT105zjfl0IjM7KQI" alt=""><figcaption></figcaption></figure>

After getting the function ID, we then registers it with the Autogen AI agents to enable the assistant to call the function.

```python
function = await integry.functions.get("slack-post-message", user_id)

function.register_with_autogen_agents(
    register_function,
    caller=assistant,
    executor=user_proxy,
    user_id=user_id,
)
```

## 4. **Connect Your Slack Account**

To allow the agent to send a message on Slack on your user's behalf, the user must connect their Slack account. To connect a Slack account against the provided user ID, execute the following snippet.

```python
slack = await integry.apps.get("slack", user_id)
print(slack.login_url)
```

This will print a URL which can be opened in a web browser to connect Slack.

## 5. Execute Agent

This will execute the agent and send a **Hello from autogen to the team** message in the Slack random channel<mark style="color:blue;">.</mark>

```python
chat_result = await user_proxy.a_initiate_chat(
    assistant,
    message="Say hello from autogen to my team on slack in #random channel",
)
```

This will send the message to the slack channel. Here is reference image

<figure><img src="/files/4PpItoR5Ynuh8hTSxh7H" alt=""><figcaption></figcaption></figure>

The message has been sent successfully in slack #random channel. You can verify the successful message delivery by checking the highlighted content in the response below.

<pre class="language-json"><code class="lang-json">User (to Assistant):

Say hello from autogen to my team on slack in #random channel

--------------------------------------------------------------------------------

>>>>>>>> USING AUTO REPLY...

>>>>>>>> USING AUTO REPLY...
Assistant (to User):

***** Suggested tool call (call_ClWxh8HVMTaR74K91HI3ixEa): slack-post-message *****
Arguments: 
{"input":{"channel":"#random","text":"Hello from autogen to the team!"}}
***********************************************************************************

--------------------------------------------------------------------------------

>>>>>>>> EXECUTING ASYNC FUNCTION slack-post-message...
User (to Assistant):

***** Response from calling tool (call_ClWxh8HVMTaR74K91HI3ixEa) *****
<strong><a data-footnote-ref href="#user-content-fn-1">{</a>"<a data-footnote-ref href="#user-content-fn-1">network_code":200,"output":{"</a>ok":true,"channel":"C086HKZMTSS","ts":"1735395154.257879","message":{"user":"U086GBQHLG0","type":"message","ts":"1735395154.257879","bot_id":"B086E311JTB","app_id":"A6FQL4KQC","text":"Hello from autogen to the team!","team":"T086682UW57","bot_profile":{"id":"B086E311JTB","app_id":"A6FQL4KQC","name":"Integry","icons":{"image_36":"https://slack-files2.s3-us-west-2.amazonaws.com/avatars/2017-08-09/225182834294_8020ddc74d7822b48ea1_36.png","image_48":"https://slack-files2.s3-us-west-2.amazonaws.com/avatars/2017-08-09/225182834294_8020ddc74d7822b48ea1_48.png","image_72":"https://slack-files2.s3-us-west-2.amazonaws.com/avatars/2017-08-09/225182834294_8020ddc74d7822b48ea1_72.png"},"deleted":false,"updated":1734709233,"team_id":"T086682UW57"},"blocks":[{"type":"rich_text","block_id":"BjV","elements":[{"type":"rich_text_section","elements":[{"type":"text","text":"Hello from autogen to the team!"}]}]}]}}}
</strong>**********************************************************************

--------------------------------------------------------------------------------

>>>>>>>> USING AUTO REPLY...

>>>>>>>> USING AUTO REPLY...
Assistant (to User):

<strong>I've posted the message "Hello from autogen to the team!" to the #random channel on Slack. TERMINATE
</strong>
--------------------------------------------------------------------------------

</code></pre>

[^1]: This success response show's that message has been sent successfully in slack channel.


---

# Agent Instructions: 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:

```
GET https://docs.integry.ai/functions/agent-frameworks/autogen.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
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.