> For the complete documentation index, see [llms.txt](https://guide.tiledesk.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://guide.tiledesk.com/tiledesk-mcp-tools/tiledesk-communicator-mcp.md).

# Tiledesk Communicator MCP

The **Tiledesk Communicator MCP** gives an AI Agent direct control over conversation actions inside Tiledesk. Instead of adding separate workflow actions for replies, waiting, human handoff, availability checks, attributes, and conversation closing, you can describe the logic in the AI Prompt and let the Agent execute the required tools automatically.

With the **Communicator MCP**, an AI Agent can ask users for information, store conversation data, send replies, check whether human agents are available, transfer conversations, and close them when the interaction is complete.

<figure><img src="https://1328774413-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FVguk6y5OMn3pAaHZSiwU%2Fuploads%2FgIcgTG9SKcUXNOaCIbS1%2FGuide%20Articles.png?alt=media&amp;token=39acff67-d03e-444e-9682-8b0edae0e1c3" alt=""><figcaption></figcaption></figure>

### Why use the Communicator MCP

Normally, conversational flows require several separate actions to reply to the user, wait for input, check agent availability, transfer to a human, store values, or close the conversation.

With the **Tiledesk Communicator MCP**, these actions become tools available directly to the AI Agent.

Instead of building every step manually in Design Studio, you can describe the desired behavior in the AI Prompt, for example:

“Ask the user about the issue. If it is resolved, close the conversation. If not, check whether a human agent is available and transfer the conversation.”

The AI Agent can then decide which Communicator tools to use and execute the flow automatically.

This makes it possible to create multi-step conversational workflows with much less manual flow configuration.

### Available tools

* **REPLY\_TO\_USER** — sends a message to the user without waiting for a response.
* **REPLY\_AND\_AWAIT\_RESPONSE** — sends a question and waits for the user's next message.
* **SET\_ATTRIBUTES** — stores information collected during the conversation.
* **CHECK\_AVAILABLE\_AGENTS** — checks whether human agents are currently available.
* **TRANSFER\_TO\_AGENT** — transfers the active conversation to a human agent.
* **CLOSE\_CONVERSATION** — closes the conversation when the interaction is complete.

<figure><img src="https://1328774413-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FVguk6y5OMn3pAaHZSiwU%2Fuploads%2Fw7QGJMOs2aqcs2OzpM6J%2FUntitled.001.001.png?alt=media&amp;token=a6848bf8-fa38-4d23-a427-4bc373aa781c" alt=""><figcaption></figcaption></figure>

### How to use Tiledesk Communicator MCP

{% stepper %}
{% step %}

#### Enable the Communicator MCP

Open the AI Prompt action where you want to manage the conversation.

Enable the **Tiledesk Communicator MCP** so the AI Agent can access its tools.

<figure><img src="https://1328774413-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FVguk6y5OMn3pAaHZSiwU%2Fuploads%2Ft4nrZonqmKt80EhZ3Y34%2F66.png?alt=media&amp;token=5d8d18de-58a0-4814-ad81-98d5ce294486" alt=""><figcaption></figcaption></figure>

<figure><img src="https://1328774413-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FVguk6y5OMn3pAaHZSiwU%2Fuploads%2F2KtU4BFpeMGGu9ICa4e7%2F67.png?alt=media&amp;token=728ad541-15d2-4607-8797-91a43e3cbdfe" alt=""><figcaption></figcaption></figure>

<figure><img src="https://1328774413-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FVguk6y5OMn3pAaHZSiwU%2Fuploads%2FZaxFQ9aaR78rQqBies3j%2F68.png?alt=media&amp;token=6edf6522-6b39-4d8e-8276-f1bf136b4a7f" alt=""><figcaption></figcaption></figure>
{% endstep %}

{% step %}

#### Add a simple prompt

In this example, the AI Agent asks the user about the issue, stores it, and offers human support.

**Prompt example:**

You are a Tiledesk support AI Agent.

Follow this flow:

1. Welcome the user and ask:\
   “Hi! How can I help you today? Please briefly describe your issue.”
2. Use `SET_ATTRIBUTES` to save `support_issue`.
3. Confirm that the request has been recorded.
4. Ask whether the user wants to speak with a human agent.
5. If yes, use `CHECK_AVAILABLE_AGENTS` and then `TRANSFER_TO_AGENT`.
6. If no, send a closing message and use `CLOSE_CONVERSATION`.

<figure><img src="https://1328774413-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FVguk6y5OMn3pAaHZSiwU%2Fuploads%2Frvt9A0Os2zfviPVEUnMk%2FGuide%20Articles%20(1).png?alt=media&amp;token=46587a26-bb8d-4f2c-9d50-6be77b58e1ba" alt=""><figcaption></figcaption></figure>
{% endstep %}

{% step %}

### Test the flow

Start a new conversation and enter a simple request, for example:

> I can't access my account.

The AI Agent should ask about the issue, store the information using **SET\_ATTRIBUTES**, and continue the conversation.

If the user requests human support, the Agent can check availability with **CHECK\_AVAILABLE\_AGENTS** and then use **TRANSFER\_TO\_AGENT**.

If the conversation is complete, the Agent can use **CLOSE\_CONVERSATION**.

<figure><img src="https://1328774413-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FVguk6y5OMn3pAaHZSiwU%2Fuploads%2FKwdLHoYSJBakI0bZ3sQz%2FGuide%20Articles%20(2).png?alt=media&amp;token=8bc5f7e8-c11b-4861-a261-e2096df223a5" alt=""><figcaption></figcaption></figure>
{% endstep %}
{% endstepper %}

### Using attributes

Use **SET\_ATTRIBUTES** when you want to save information collected during the conversation.

For example:

```
support_issue = Account access problem
```

The value can then be reused in later Design Studio blocks:

```
{{support_issue}}
```

### Example use cases

* Customer support with human handoff
* Lead qualification
* Request classification and triage
* Progressive information collection
* Conversation closing

### Next step

The Communicator MCP can also be combined with other Tiledesk MCP tools.

For example, you can use the [Tiledesk Knowledge Base MCP](/tiledesk-mcp-tools/tiledesk-knowledge-base-mcp.md) to retrieve answers and the [Tiledesk Data Tables MCP](/tiledesk-mcp-tools/tiledesk-data-tables-mcp.md) to save structured information such as leads or requests.


---

# 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 by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://guide.tiledesk.com/tiledesk-mcp-tools/tiledesk-communicator-mcp.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

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.