notify-completion-mcp-server

by nakat-t

2 126 downloads Not rated yet MIT
GitHub

About

The MCP server that provides notifications when AI agents complete long-running tasks. This allows you to leave your desktop without forgetting about tasks.

Details

License
MIT

Explore

- Provides notifications when AI agents complete long-running tasks
- Offers two MCP tools: start-timer and check-overdue
- Configurable overdue command for custom notifications
- Supports adjustable threshold to skip short tasks
- Works with Cline, Claude Desktop, Cursor, and GitHub Copilot

Setting up with Highlight

This MCP is not yet compatible with Highlight’s one-click setup. However, you can still use it with Highlight by following these steps:

  1. Download and install Highlight from highlightai.com/download
  2. Navigate to the plugins tab and select "Add Custom Plugin"
  3. Configure the plugin with the settings below
    Plugin Name notify-completion-mcp-server
    Command (node, npx, python, etc.)

    Please refer to the README for specific instructions on how to obtain API keys or other required environment variables.

  4. Enable "Start Automatically" if you want the plugin to start when Highlight launches

From the repository

1. Add config to the client (If using Cline, edit cline_mcp_settings.json):

{
  "mcpServers": {
    "notify-completion": {
      "command": "npx",
      "args": ["-y", "@nakat-t/notify-completion-mcp-server@latest"],
      "env": {
        "OVERDUE_COMMAND": "<Command to execute very flashy effects>",
      },
      "autoApprove": [ "start-timer", "check-overdue" ]
    }
  }
}

2. Change the config environment variables to your value.
OVERDUE_COMMAND: Write a command to notify you when the task is complete.
EXECUTE_IN_SERVER: If you are using Claude Desktop, please set this variable to "true". When using AI capable of executing commands on their own, there is no need to define this variable.

3. Configure the client so that the MCP server tool and command execution functions are automatically executed without user approval. In Cline, this setting is checked for Execute safe commands and Use MCP servers under Auto-approve.

4. Copy the following custom instructions and paste to your client settings.

- THRESHOLD: If the task execution time is less than the specified number of seconds, the overdue command is not executed. Default is 0. (always executed)

- OVERDUE_COMMAND: Specify the command to execute when the task is completed.

- EXECUTE_IN_SERVER: If this variable is defined, the overdue command is executed within the MCP server instead of being executed by the AI agent.

startTime

Start time returned by start-timer

1. start-timer: Start a timer to measure task execution time.
Output: {"startTime": "<Current time in ISO format>"}
Description: We instruct AI agents to call this tool first before executing tasks. Use custom instruction to do this.
2. check-overdue: Check if the task is overdue based on the given threshold.
Required inputs:
startTime: Start time returned by start-timer
Output:
If the task execution time is not overdue: {"elapsed": "<seconds>", "units": "seconds", "overdue": false}
* If the task execution time is overdue: {"elapsed": "<seconds>", "units": "seconds", "overdue": true, "command": "${OVERDUE_COMMAND}"}

Claude Desktop / Cursor

Paste into your MCP client config file to install this server.

{
    "mcpServers": {
        "notify-completion-mcp-server": {
            "notify-completion": {
                "command": "npx",
                "args": [
                    "-y",
                    "@nakat-t/notify-completion-mcp-server@latest"
                ],
                "env": {
                    "OVERDUE_COMMAND": "<Command to execute very flashy effects>"
                },
                "autoApprove": [
                    "start-timer",
                    "check-overdue"
                ]
            }
        }
    }
}

McpServers

{
    "notify-completion": {
        "command": "npx",
        "args": [
            "-y",
            "@nakat-t/notify-completion-mcp-server@latest"
        ],
        "env": {
            "OVERDUE_COMMAND": "<Command to execute very flashy effects>"
        },
        "autoApprove": [
            "start-timer",
            "check-overdue"
        ]
    }
}

npm version badge
cline:worked badge
claude desktop:worked badge
cursor:worked badge
github copilot:worked badge

The MCP server that provides notifications when AI agents complete long-running tasks. This allows you to leave your desktop without forgetting about tasks.

How It Works

How It Works - mermaid diagram

The notify-completion-mcp-server addresses the issue of users needing to actively monitor their computer for AI task completion. By providing a notification mechanism triggered when tasks are finished, users can multitask more efficiently without repeatedly checking task status.

Installation

1. Add config to the client (If using Cline, edit cline_mcp_settings.json):

{
  "mcpServers": {
    "notify-completion": {
      "command": "npx",
      "args": ["-y", "@nakat-t/notify-completion-mcp-server@latest"],
      "env": {
        "OVERDUE_COMMAND": "<Command to execute very flashy effects>",
      },
      "autoApprove": [ "start-timer", "check-overdue" ]
    }
  }
}

2. Change the config environment variables to your value.
OVERDUE_COMMAND: Write a command to notify you when the task is complete.
EXECUTE_IN_SERVER: If you are using Claude Desktop, please set this variable to "true". When using AI capable of executing commands on their own, there is no need to define this variable.

3. Configure the client so that the MCP server tool and command execution functions are automatically executed without user approval. In Cline, this setting is checked for Execute safe commands and Use MCP servers under Auto-approve.

4. Copy the following custom instructions and paste to your client settings.

Custom Instructions

```markdown

No reviews yet — be the first

Sign in to leave a review

Use Google, GitHub, or an email account so ratings stay tied to real people.

Email sign in

No reviews posted yet.