klbr/n8nworkflows.xyz
1
0
Fork 0
mirror of https://github.com/khoaliber/n8nworkflows.xyz.git synced 2026-04-28 00:29:22 +00:00
Code Issues Packages Projects Releases Wiki Activity
Files
8a7e37ebc1ed6ebfac4cfe5546c77c9309eaafec
n8nworkflows.xyz/workflows/AI-Powered Gmail MCP Server-3623/readme-3623.md
T
nusquama 32d6fb14a7 creation
2025-11-12 20:11:26 +01:00

15 KiB
Raw Blame History

AI-Powered Gmail MCP Server

https://n8nworkflows.xyz/workflows/ai-powered-gmail-mcp-server-3623

AI-Powered Gmail MCP Server

1. Workflow Overview

This workflow, titled AI-Powered Gmail MCP Server, is designed to automate Gmail email operations by integrating an external AI Model Control Plane (MCP) Server with native Gmail nodes in n8n (version 1.88.0+). It targets developers, automation engineers, and power users who want to leverage AI-generated content for composing, replying, fetching, and managing email conversations in Gmail with minimal manual intervention.

The workflow is logically divided into the following blocks:

  • 1.1 MCP Trigger Input Reception: Receives standardized HTTP requests from the external MCP Server containing AI-generated content and email metadata.
  • 1.2 Gmail Email Actions: Executes Gmail operations such as sending new emails, replying to threads, fetching email details, and sending emails with wait-for-reply functionality, all driven by AI-generated inputs from the MCP trigger.

2. Block-by-Block Analysis

1.1 MCP Trigger Input Reception

  • Overview:
    This block serves as the entry point for the workflow. It listens for incoming HTTP requests from the MCP Server, which contain AI-generated email content and metadata. It standardizes these inputs for downstream Gmail nodes.

  • Nodes Involved:

    • MCP_GMAIL

  • Node Details:

    • MCP_GMAIL
      • Type & Role: MCP Trigger node (@n8n/n8n-nodes-langchain.mcpTrigger), specialized for receiving AI Model Control Plane HTTP calls.
      • Configuration:
        • HTTP path: /mcp/:tool/gmail (dynamic segment :tool allows flexibility for different AI tools).
        • Listens for incoming HTTP POST requests from the MCP Server.
      • Expressions/Variables:
        • Dynamic URL parameter :tool is available for routing or logging if needed.
      • Input/Output:
        • No input connections (trigger node).
        • Outputs standardized AI-generated data fields such as To, Subject, Message, Message_ID, Simplify (boolean), etc., to connected Gmail nodes.
      • Version Requirements:
        • Requires n8n version 1.88.0+ due to MCP Trigger node availability.
      • Potential Failures:
        • Authentication failures if MCP Server is not properly authorized to call the webhook.
        • Malformed or incomplete HTTP payloads causing downstream nodes to fail.
        • Network or webhook URL misconfiguration.
      • Sub-workflow:
        • None.

1.2 Gmail Email Actions

  • Overview:
    This block performs Gmail operations using native Gmail nodes, all driven by AI-generated content received from the MCP trigger. It supports sending new emails, replying to existing threads, fetching email content, and sending emails with a wait-for-reply mechanism.

  • Nodes Involved:

    • SEND_EMAIL
    • REPLY_EMAIL
    • GET_EMAIL
    • SEND_AND_WAIT

  • Node Details:

    • SEND_EMAIL

      • Type & Role: Gmail Tool node (v2.1), sends new emails.
      • Configuration:
        • sendTo: Populated dynamically from MCP input field To.
        • subject: Populated dynamically from MCP input field Subject.
        • message: Populated dynamically from MCP input field Message.
        • descriptionType: Manual (content is explicitly set).
        • No additional options configured.
      • Expressions/Variables:
        • Uses $fromAI() expression to override fields with AI-generated data from MCP trigger.
      • Input/Output:
        • Input: Connected from MCP_GMAIL node output (ai_tool output).
        • Output: Standard Gmail send result (message metadata).
      • Version Requirements:
        • Requires Gmail Tool node v2.1 and n8n v1.88.0+.
      • Potential Failures:
        • Gmail OAuth2 authentication errors.
        • Invalid email addresses or malformed message content.
        • API rate limits or quota exceeded.
      • Sub-workflow:
        • None.

    • REPLY_EMAIL

      • Type & Role: Gmail Tool node (v2.1), replies to existing email threads.
      • Configuration:
        • messageId: Populated dynamically from MCP input field Message_ID to identify the email to reply to.
        • message: AI-generated reply content from MCP input field Message.
        • operation: Set to reply.
        • descriptionType: Manual.
      • Expressions/Variables:
        • Uses $fromAI() to dynamically override message and messageId.
      • Input/Output:
        • Input: Connected from MCP_GMAIL node output (ai_tool output).
        • Output: Reply message metadata.
      • Version Requirements:
        • Gmail Tool v2.1, n8n v1.88.0+.
      • Potential Failures:
        • Missing or invalid Message_ID causing failure to locate thread.
        • Authentication or API errors.
        • Reply content formatting issues.
      • Sub-workflow:
        • None.

    • GET_EMAIL

      • Type & Role: Gmail Tool node (v2.1), fetches email details for a given message ID.
      • Configuration:
        • messageId: From MCP input field Message_ID.
        • simple: Boolean flag from MCP input Simplify to control output verbosity.
        • operation: Set to get.
      • Expressions/Variables:
        • Uses $fromAI() for both messageId and simple.
      • Input/Output:
        • Input: Connected from MCP_GMAIL node output (ai_tool output).
        • Output: Email message data (full or simplified).
      • Version Requirements:
        • Gmail Tool v2.1, n8n v1.88.0+.
      • Potential Failures:
        • Invalid or missing Message_ID.
        • API errors or permission issues.
      • Sub-workflow:
        • None.

    • SEND_AND_WAIT

      • Type & Role: Gmail Tool node (v2.1), sends an email and pauses workflow until a reply is received in the same thread.
      • Configuration:
        • sendTo, subject, message: All dynamically populated from MCP input fields.
        • operation: sendAndWait.
        • responseType: freeText (expects free-text reply content).
        • No additional options set.
      • Expressions/Variables:
        • Uses $fromAI() for sendTo, subject, and message.
      • Input/Output:
        • Input: Connected from MCP_GMAIL node output (ai_tool output).
        • Output: Data from the reply message received after waiting.
      • Version Requirements:
        • Gmail Tool v2.1, n8n v1.88.0+.
      • Potential Failures:
        • Timeout if no reply is received within configured wait period (default or custom).
        • Authentication or API errors.
        • Invalid recipient or message content.
      • Sub-workflow:
        • None.

3. Summary Table

Node Name Node Type Functional Role Input Node(s) Output Node(s) Sticky Note
MCP_GMAIL MCP Trigger (mcpTrigger) Receives AI-generated inputs from MCP Server None SEND_EMAIL, REPLY_EMAIL, GET_EMAIL, SEND_AND_WAIT Entry point webhook for MCP Server calls. Path: /mcp/:tool/gmail. Requires n8n v1.88.0+.
SEND_EMAIL Gmail Tool (v2.1) Sends new emails with AI-generated content MCP_GMAIL None Sends messages via Google API. Uses AI overrides for To, Subject, Message fields.
REPLY_EMAIL Gmail Tool (v2.1) Replies to existing email threads using AI content MCP_GMAIL None Replies to threads using Message ID and AI-generated reply content.
GET_EMAIL Gmail Tool (v2.1) Fetches email details for analysis or processing MCP_GMAIL None Retrieves message data by Message ID. Supports simplified output option.
SEND_AND_WAIT Gmail Tool (v2.1) Sends email and waits for reply in thread MCP_GMAIL None Sends email and pauses workflow until reply received. Outputs reply data.

4. Reproducing the Workflow from Scratch

  1. Create MCP_GMAIL Node (Trigger):

    • Add node of type MCP Trigger (@n8n/n8n-nodes-langchain.mcpTrigger).
    • Set HTTP path to /mcp/:tool/gmail.
    • This node will receive HTTP POST requests from your MCP Server containing AI-generated email data.
    • No credentials needed here unless your MCP Server requires webhook authentication (configure accordingly).

  2. Create SEND_EMAIL Node:

    • Add node of type Gmail Tool (version 2.1).
    • Set operation to send (default).
    • Configure fields:
      • sendTo: Use expression ={{ $fromAI('To', '', 'string') }} to dynamically receive recipient email from MCP input.
      • subject: Use expression ={{ $fromAI('Subject', '', 'string') }}.
      • message: Use expression ={{ $fromAI('Message', '', 'string') }}.
      • descriptionType: Set to manual.
    • Under credentials, select your configured Gmail OAuth2 credential.
    • Connect input from MCP_GMAIL node output (ai_tool output).

  3. Create REPLY_EMAIL Node:

    • Add node of type Gmail Tool (v2.1).
    • Set operation to reply.
    • Configure fields:
      • messageId: Expression ={{ $fromAI('Message_ID', '', 'string') }} to identify email thread.
      • message: Expression ={{ $fromAI('Message', '', 'string') }} for AI-generated reply content.
      • descriptionType: manual.
    • Use the same Gmail OAuth2 credentials as above.
    • Connect input from MCP_GMAIL node output (ai_tool output).

  4. Create GET_EMAIL Node:

    • Add node of type Gmail Tool (v2.1).
    • Set operation to get.
    • Configure fields:
      • messageId: Expression ={{ $fromAI('Message_ID', '', 'string') }}.
      • simple: Expression ={{ $fromAI('Simplify', '', 'boolean') }} to toggle simplified output.
    • Use Gmail OAuth2 credentials.
    • Connect input from MCP_GMAIL node output (ai_tool output).

  5. Create SEND_AND_WAIT Node:

    • Add node of type Gmail Tool (v2.1).
    • Set operation to sendAndWait.
    • Configure fields:
      • sendTo: Expression ={{ $fromAI('To', '', 'string') }}.
      • subject: Expression ={{ $fromAI('Subject', '', 'string') }}.
      • message: Expression ={{ $fromAI('Message', '', 'string') }}.
      • responseType: Set to freeText.
    • Use Gmail OAuth2 credentials.
    • Connect input from MCP_GMAIL node output (ai_tool output).

  6. Connect Nodes:

    • All Gmail nodes (SEND_EMAIL, REPLY_EMAIL, GET_EMAIL, SEND_AND_WAIT) receive input from the MCP_GMAIL node’s ai_tool output.
    • No further connections are required unless extending the workflow.

  7. Credential Setup:

    • Configure Gmail OAuth2 credentials in n8n under Settings > Credentials > New > Google > Gmail (OAuth2 API).
    • Ensure the Gmail API is enabled in your Google Cloud project.
    • Configure MCP Server authentication as needed (e.g., Header Auth credential in n8n or external).

  8. Activate Workflow:

    • Set the workflow toggle to active.
    • Copy the webhook URL from the MCP_GMAIL node and configure your MCP Server to send HTTP requests to this URL with the required AI-generated payloads.

5. General Notes & Resources

Note Content Context or Link
This workflow requires n8n version 1.88.0 or higher due to the usage of native Gmail Tool nodes and MCP Trigger node. n8n official documentation: https://docs.n8n.io/
Gmail OAuth2 credentials must be configured with the Gmail API enabled in Google Cloud Console. Google Cloud Console: https://console.cloud.google.com/apis/library/gmail.googleapis.com
The MCP Server must be configured to call the n8n webhook URL with properly formatted AI-generated email data fields such as To, Subject, Message, Message_ID, and Simplify. Workflow description and MCP Server documentation (external)
Use the $fromAI() expression to dynamically override node parameters with AI-generated content received via the MCP trigger. n8n expression documentation: https://docs.n8n.io/nodes/expressions/
For error handling, consider connecting the error outputs of Gmail nodes to an Error Trigger node to log or notify on failures. n8n error handling best practices.
Customize AI prompts and post-response actions by extending the workflow after the Gmail nodes, e.g., logging to databases or updating CRM systems. Workflow customization suggestions in description.
The Gmail Tool nodes support advanced search filters and options; customize the GET_EMAIL node’s search parameters if needed to target specific emails. Gmail search operators: https://support.google.com/mail/answer/7190?hl=en

This documentation provides a detailed, structured reference to understand, reproduce, and extend the AI-Powered Gmail MCP Server workflow in n8n, ensuring robust integration between AI-generated content and Gmail automation.

Reference in New Issue View Git Blame Copy Permalink