DavidFarago/vise-logger

Vise Logger

Vise Logger is an MCP server that captures, rates, and archives Vise Coding sessions from various IDEs, storing them for future reference, comparisons, and searchability.

Installation and Setup

Local installation

This project uses uv for package and environment management.

1. Create a virtual environment:

   uv venv
   

2. Activate the virtual environment:

   • On macOS/Linux:

     source .venv/bin/activate
     

   • On Windows:

     .venv\Scripts\activate
     

3. Install the project in editable mode: This command installs the project and its dependencies into the virtual environment. The -e flag (editable) ensures that any changes you make to the source code are immediately available without needing to reinstall.

   uv sync
   

   This installs the runtime dependencies, in editable mode. To additionally install the dev dependencies, add argument --extra dev

Installation of MCP Server via PyPI

Making use of the MCP server in your AI coding tool

In Cline and Roo:

    "vise-logger": {
      "disabled": false,
      "timeout": 60,
      "type": "stdio",
      "command": "uv",
      "args": [
        "--directory",
        "/home/emergency/git/vise-logger",
        "run",
        "vise-logger"
      ]
    }

Running the MCP Server

To run the MCP server directly from the command line:

vise-logger

MCP Tools

This server provides two tools:

rate_and_upload(stars: float, comment: str = "")

Rates and archives the current Vise Coding session. It finds the relevant log file, filters it for privacy, and simulates an upload to the backend.

configure_log_dir()

Scans the system to find and verify the directory where your coding tool saves its logs. This helps speed up future searches.

Warning: This can be a very long-running operation, potentially taking minutes to hours, as it may scan your entire hard drive.

Running Tests

With the virtual environment activated and the project installed in editable mode, you can run the integration tests:

python3 -m unittest discover tests

Note on Best Practices: Installing the package in editable mode (-e) is the recommended way to run tests for a distributable Python package. It correctly resolves imports without needing to modify sys.path, which is a less robust method. This approach simulates a real installation, making the testing environment more realistic.

REST Endpoints The Sessions Are Sent to

The endpoint should have Content-Type: multipart/form-data and the following form fields:

• file (required): The zip file containing the AI coding session data
• metadata (required): JSON string with AI coding session metadata. The metadata JSON structure:

json{
  "marker": "string (required)",
  "tool": "string (required)",
  "stars": "number (required)",
  "comment": "string (optional)"
}

The official web application for Vise Logger is www.viselo.gr. You can test it via

curl -X POST \
  -F "file=@session.zip" \
  -F 'metadata={"marker": "Rated session at 2025-07-30-20-56-11: 1.9 stars.", "tool": "curl", "stars": 1.9, "comment": "Just a curl test"}' \
  "https://studio--viselog.us-central1.hosted.app/api/v1/sessions"

Debugging

MCP-Server's log file: ~/.local/state/vise-logger/log/mcp_server.log (path depends on your machine, see https://pypi.org/project/platformdirs/)