MCP Meeting Summary System Server

Provides AI-driven meeting summaries and Q&A using MCP tools with OpenAI and Selenium-based processing.

python

GitHub Stars

python

Language

7 months ago

First Indexed

3 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "w87878-advanced_mcp": {
      "command": "uv",
      "args": [
        "run",
        "./mcp_server.py"
      ],
      "env": {
        "DIR": "/workspace",
        "MODEL": "gpt-4",
        "MCP_MODE": "streamable-http",
        "MAX_TOKENS": "2048",
        "TEMPERATURE": "0.7",
        "USER_DATA_DIR": "/path/to/data",
        "MAX_ITERATIONS": "10",
        "OPENAI_API_KEY": "sk-...",
        "REQUEST_TIMEOUT": "60",
        "MAX_INPUT_LENGTH": "4096",
        "PROFILE_DIRECTORY": "/path/to/profiles"
      }
    }
  }
}

You can run a complete MCP server setup that combines an autonomous meeting summarizer with a Q&A interface. This system uses a tool-oriented MCP architecture to generate high-quality Markdown meeting reports from transcripts and enables you to ask questions and receive concise, structured answers.

How to use

Start the MCP tool server and the agent interface locally, then connect your client to run queries against transcripts or generated summaries. The tool server handles the summarization workflow, while the agent interface provides an interactive question-answering experience. Use a compatible MCP client to send your questions or paste your meeting transcripts, and you will receive a structured Markdown summary or an answer with key points.

How to install

Prerequisites you need before installing: a Python environment for backend services, Node.js and npm for frontend tooling, and a modern shell to run commands.

# Install Python dependencies for the backend tools (as needed)
uv pip install -r pyproject.toml
# Install frontend dependencies
cd front
npm install

# Run the MCP tool server (stdio mode) and the agent interface (stdio mode)
cd api
uv run ./mcp_server.py

# In a separate session, start the MCP agent interface
uv run ./main.py

Configuration and environment

Create an environment file to provide API keys and runtime options. The following variables control API access, runtime modes, and resource limits. Use your own values where indicated.

OPENAI_API_KEY=your_key_here
MCP_MODE=streamable-http, sse, stdio
MAX_ITERATIONS=your_max_iterations_here
MAX_TOKENS=your_max_tokens_here
TEMPERATURE=your_temperature_here
REQUEST_TIMEOUT=your_timeout_here
MODEL=your_model_here
USER_DATA_DIR=your_data_dir_here
PROFILE_DIRECTORY=your_profile_directory_here
DIR=your_directory_here
MAX_INPUT_LENGTH=your_max_length_here

Available tools

summarize_meeting

MCP tool that generates meeting summaries by coordinating OpenAI processing with Selenium-driven workflows to produce structured Markdown reports.

markdown_reformatter

Converts OpenAI-generated content into a structured Markdown meeting minutes format for easy sharing and archiving.

Built by

VeilStrat

AI signals for GTM teams