home / mcp / gemini daily mcp server

Gemini Daily MCP Server

Provides token-efficient access to the Gemini CLI with session-aware execution and on-demand help for full documentation.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "kdoooh000h-gemini-daily-mcp-compressed": {
      "command": "node",
      "args": [
        "/path/to/gemini-daily-mcp-compressed/dist/index.js"
      ],
      "env": {
        "LOG_LEVEL": "info",
        "MAX_SESSIONS": "100",
        "GEMINI_ENABLED": "true",
        "SESSION_TTL_HOURS": "24",
        "GEMINI_CLI_COMMAND": "gemini"
      }
    }
  }
}

You deploy the Gemini Compressed MCP Server to run the Gemini CLI with token-efficient MCP schemas. It delivers full Gemini CLI functionality while reducing tool schema size through Progressive Disclosure, helping you manage many conversations and tools at lower cost.

How to use

You interact with the MCP server through your MCP client by selecting the gemini_compressed server as the endpoint. You can run Gemini commands, manage conversational sessions, switch models, and reset sessions as needed. The help tool will provide comprehensive documentation when you need it, while most interactions stay lean to save tokens.

How to install

Prerequisites checklists and steps are below. Follow them in order to get your MCP server up and running.

# Prerequisites
# - Node.js installed
# - npm available

# Optional: ensure Gemini CLI is installed and configured on your system separately
# - gemini auth login

# Install the MCP server package
npm install

# Build (if a build step is provided in your workflow)
npm run build

# Start the MCP server using the stdio configuration shown in the metadata
# The runtime command is defined in the config below

Configuration and runtime details

The MCP server is exposed as a stdio (local) server. You run it by launching Node with the generated entry point, and you provide the environment for Gemini integration as shown.

{
  "mcpServers": {
    "gemini_compressed": {
      "command": "node",
      "args": ["/path/to/gemini-daily-mcp-compressed/dist/index.js"],
      "env": {
        "GEMINI_ENABLED": "true",
        "SESSION_TTL_HOURS": "24",
        "MAX_SESSIONS": "100",
        "GEMINI_CLI_COMMAND": "gemini",
        "LOG_LEVEL": "info"
      }
    }
  }
}

Notes on usage patterns

- Use sessions to persist context across multi-turn conversations and save tokens. A session can significantly reduce token usage in long interactions. - You can specify the Gemini model when needed to handle more complex analyses. - If you need to reset a session, you can do so to start fresh while keeping the rest of the workflow intact.

Troubleshooting

If you encounter connection issues, verify that the MCP server process starts correctly and that the environment variables are set as shown. Use the ping tool to test connectivity and the help tool to pull up detailed guidance.

Available tools

gemini

Execute Gemini CLI with optional session support, including model selection and session reset for multi-turn conversations.

listSessions

List all active conversation sessions to manage context and reminders.

ping

Test the MCP server connectivity and availability.

help

Provide enhanced, comprehensive documentation for all available tools on demand.