> For the complete documentation index, see [llms.txt](https://docs.diaflow.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.diaflow.io/chat/connectors.md).

# Connectors

### Overview

**Connectors** allow Diaflow to securely interact with third-party tools, databases, messaging platforms, and cloud services. By enabling connectors, your AI models and Chat agents gain the ability to fetch real-time context, read documents, create database records, send notifications, and perform automated actions on your behalf.

Diaflow supports two primary types of connectors:

1. **Pre-built Library Connectors**: Native, one-click integrations with popular tools (Google Workspace, Slack, Airtable, HubSpot, Salesforce, Linear, etc.).
2. **Custom MCP (Model Context Protocol) Servers**: Connect any remote MCP server to extend your agents with custom internal APIs and tools.

***

### Workspace vs. Personal Connectors

At the top of the Connectors management screen, you can switch between two access scopes:

* **Personal Connectors**: Connected with your individual credentials. Tools and actions authorized here are private to your user account and cannot be accessed or triggered by other team members.
* **Workspace Connectors**: Shared across your entire organization or workspace. Enables shared team tools (such as corporate CRM, shared Google Drive, or shared Slack bots) so all authorized teammates can utilize them in workflows and chat sessions.

<figure><img src="https://3864624312-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FRdo1pkMnofwxf3lpWftw%2Fuploads%2Fd3opU0wv38PUBMo0LihI%2Frounded-image%20(9).png?alt=media&amp;token=04ba3478-1544-4081-bd23-b82e22733aff" alt=""><figcaption></figcaption></figure>

***

### Adding a Connector from the Library

Diaflow provides an extensive library of pre-built integrations for CRM, communication, cloud storage, project management, and developer tools.

<figure><img src="https://3864624312-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FRdo1pkMnofwxf3lpWftw%2Fuploads%2FQbmDvsERbJSD0ZphrnPW%2Frounded-image%20(11).png?alt=media&amp;token=7e42db40-3f84-43da-8974-6f9c80101bcb" alt=""><figcaption></figcaption></figure>

#### Step-by-step Guide

1. Navigate to your **Settings** and select **Connectors** from the left sidebar.
2. In the top-right corner, click the **Add** button and select **Library**.
3. Use the search bar or browse the list of available integrations:
   * **Communication & Collaboration**: Slack, Microsoft Teams, Telegram, WhatsApp Business, Gmail, Microsoft Outlook.
   * **Productivity & Docs**: Google Drive, Google Docs, Google Sheets, Google Photos, Airtable, Microsoft Outlook Calendar.
   * **CRM & Marketing**: HubSpot, Salesforce, SuiteDash, Mailchimp, Twilio SendGrid.
   * **Finance & Accounting**: QuickBooks, Xero Accounting.
   * **Project Management & Development**: Linear, AWS, Wordpress.org.
   * **Data & Search**: SerpApi, FireCrawl.
4. Click the **`+`** icon next to the service you want to connect.
5. Follow the authorization prompt (OAuth2 login or API Key input) to grant Diaflow the necessary permissions.
6. Once authenticated, the platform will immediately appear under your **Connected** list with an active status.

***

### Adding a Custom MCP Connection

Diaflow embraces Anthropic's **Model Context Protocol (MCP)**, an open standard that allows AI agents to securely connect to arbitrary external data sources and execution environments.

<figure><img src="https://3864624312-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FRdo1pkMnofwxf3lpWftw%2Fuploads%2FVwrpQEcHpR9CcZ8Eywjc%2Frounded-image%20(10).png?alt=media&amp;token=bd3e0282-dfc5-4ff9-bbff-4296003f2dc5" alt=""><figcaption></figcaption></figure>

#### When to use Custom MCP?

* When connecting internal proprietary APIs and private company databases.
* When using specialized tools built with the open-source MCP SDK.
* When running local or remote MCP servers hosted on your own infrastructure.

#### How to set up a Remote MCP Server:

1. Click the **Add** button in the top-right corner and select **Custom MCP**.
2. Fill in the connection details in the modal:
   * **Name**: A descriptive name for this connection (e.g., `Internal Database MCP`, `Stripe Analytics Server`).
   * **Remote MCP server URL**: The endpoint of your deployed MCP server (e.g., `https://mcp.yourdomain.com/` or `https://mcp.diaflow.io/service`).
3. *(Optional)* Click **Advanced settings** to configure:
   * **Authentication type**: Set to `Auto-detect (recommended)`, or specify your custom Bearer token/header authorization if required by your remote server.
4. Click **Continue** to verify the handshake and register the MCP tools with Diaflow.

> **Security Note:** Use connectors only from developers you trust. Diaflow does not manage the tools external developers offer and cannot guarantee their functionality or stability. Always adhere to the principle of least privilege when granting permissions.

***

### Managing Existing Connections

All currently authorized integrations are displayed under the **Connected** section:

* **Search**: Quickly find any connected service using the search bar.
* **Configure Settings**: Click the **Gear icon ⚙️** next to any connected platform to view permissions, re-authenticate an expired session, edit connection parameters, or remove/disconnect the service.
* **Audit & Privacy**: View how permissions are mapped and control data access under **Settings > Tools & Privacy**.