LiveKit

Provides an MCP server for semantic search across LiveKit docs and real-time web results with source-backed answers.

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": {
    "thenabilman-thenabilman_livekit_mcp_assistant": {
      "command": "python",
      "args": [
        "mcp_server_standard.py"
      ],
      "env": {
        "GROQ_API_KEY": "YOUR_GROQ_KEY",
        "TAVILY_API_KEY": "YOUR_TAVILY_KEY",
        "PINECONE_API_KEY": "YOUR_PINECONE_KEY",
        "PINECONE_INDEX_NAME": "livekit-docs"
      }
    }
  }
}

You run an MCP server that orchestrates a fast, hybrid semantic search experience for LiveKit-like documentation and data sources. It exposes a standard MCP workflow so you can query documentation or web results and receive instant, source-backed answers. This setup is designed for local development with a lightweight UI and a runnable backend, making it simple to prototype, test, and use in your own projects.

How to use

You will interact with a client that talks to the MCP server to ask questions and receive answers with sources. Start the MCP server in one terminal and the UI in another. Use the server to perform two modes: Docs mode to search local/documentation data, and Web mode to fetch real-time web results. You can then ask practical questions like how to set up the LiveKit environment and get quick, source-backed answers.

Typical usage flow:

How to install

Prerequisites you need before starting:

Conda is recommended to manage environments
Pip is used to install dependencies

Step-by-step commands to set up your environment and run the MCP server and UI:

Install and run steps

# Prerequisites
conda create -n langmcp python=3.12
conda activate langmcp
pip install -r requirements.txt

# Configure API keys in a .env file

# Example .env content
GROQ_API_KEY=your_key
TAVILY_API_KEY=your_key
PINECONE_API_KEY=your_key
PINECONE_INDEX_NAME=livekit-docs

# Start MCP Server in Terminal 1
python mcp_server_standard.py

# Start UI in Terminal 2
streamlit run app.py

App opens at http://localhost:8501

Additional sections

Configuration and environment variables you will use to run the MCP server are shown below. These keys enable core services used by the server to process queries and fetch live data.

GROQ_API_KEY=your_key
TAVILY_API_KEY=your_key
PINECONE_API_KEY=your_key
PINECONE_INDEX_NAME=livekit-docs

Files you will typically interact with include the following, which contain the MCP server, ingestion logic, and UI components:

Notes on configuration, troubleshooting, and usage

If you encounter no results, try switching to web mode or use different keywords. Slow first responses are expected on first start as the model initializes, but subsequent queries are faster due to caching and optimal search latency.

If the MCP server cannot be found, ensure you started the MCP server process with its designated command in a separate terminal.

API errors usually indicate misconfigured keys in the environment. Double-check your .env file and ensure all required keys are present and correctly named.

Key performance notes: first query typically takes 15–20 seconds while the model loads; cached queries are 2–5 seconds; search latency is under 500 ms.

Built by

VeilStrat

AI signals for GTM teams