playbooks

home / mcp / mcp chat logger mcp server

MCP Chat Logger MCP Server

📝 A FastMCP tool that automatically saves AI chat conversations to well-formatted Markdown files with timestamps and metadata.

Installation
Add the following to your MCP client configuration file.

Configuration

View docs
{
  "mcpServers": {
    "alexifeng-mcp_chat_logger": {
      "command": "uv",
      "args": [
        "--directory",
        "项目路径（例如~/MCP_Chat_Logger/）",
        "run",
        "chat_logger.py"
      ]
    }
  }
}

MCP Chat Logger is a focused tool that saves chat history as Markdown files with timestamps, organized by conversation, and stored in a configurable directory for easy viewing and sharing.

How to use

You run the MCP Chat Logger as a local MCP server and connect your MCP client to capture chat histories. When a chat session starts, the logger records each message with a time stamp and saves the conversation to a Markdown file in your chosen directory. Use a per-conversation identifier to separate different chats. Start the logger, then point your MCP client to the configured server to begin recording.

How to install

Prerequisites you need before installation:

- Python (for running the Python script chat_logger.py)

- uv (the lightweight runner used to start scripts)

Installation steps you should follow in sequence:

# 1. Clone the project repository
git clone https://github.com/yourusername/MCP_Chat_Logger.git
cd MCP_Chat_Logger

# 2. Install dependencies (if needed for Python script or runtime)
# If you use uv, ensure it is available in your environment
uv add "mcp[cli]"  # optional if you plan to use Smithery workflow

# 3. Prepare to run the logger (see next section for the runtime command)

MCP configuration and runtime workflow

You configure the local MCP server to run the logger and specify how it should be started. The following configuration uses the UV runtime to start the Python script with a designated save directory.

"chat_logger": {
  "name": "chat_logger",
  "isActive": false,
  "command": "uv",
  "args": [
    "--directory",
    "项目路径（例如~/MCP_Chat_Logger/）",
    "run",
    "chat_logger.py"
  ]
}

Available tools

saveToMarkdown

Formats and writes each chat message into a Markdown file, using a timestamp for every entry and grouping messages by session ID.

timestamp

Automatically adds a time stamp to every message to preserve chronology across sessions.

directoryConfig

Allows you to customize the save directory where all chat log files are stored.

sessionIdentification

Supports using a session/conversation ID to separate different chats within the logger output.