home / mcp / lambda capture mcp server

Lambda Capture MCP Server

Enables semantic search for macroeconomic data via remote HTTP or local MCP servers, with configurable clients.

Installation

Add the following to your MCP client configuration file.

Configuration

{
  "mcpServers": {
    "lambda-capture-mcp-server": {
      "url": "https://mcp.lambda-capture.com/v1/mcp/",
      "headers": {
        "LAMBDA_CAPTURE_API_KEY": "your_api_key"
      }
    }
  }
}

You can access Lambda Capture’s MCP server to perform semantic search over macroeconomic data. This server supports both a remote HTTP MCP endpoint and local MCP servers you run on your own machine, letting you query, list tools, and integrate into clients like Claude Desktop App. The remote endpoint is useful for quick testing, while the local options give you full control over deployment and environment.

How to use

Use an MCP client to connect to the remote HTTP endpoint or to your locally running MCP server. The remote endpoint provides a set of tools you can call to perform macroeconomic semantic searches, such as retrieving key shifts in inflation expectations or querying available tools. When you connect, your client sends a request that includes the tool you want to run and the arguments for your specific query.

To begin, choose one of the following connection options: - Remote HTTP MCP endpoint: call the endpoint to perform a semantic search against macroeconomic data. - Local MCP server: run the MCP server locally and connect to it from your client, using the local runtime endpoint. This option requires you to provide an API key in your environment.

How to install

Prerequisites depend on the runtime you choose. For MCP Typescript you need Node.js 18+ (which includes npx and npm). For MCP Python you need Python 3.11+.

Local MCP server setup steps depend on whether you use Node.js or Python. Follow the concrete steps below to install and prepare your environment.

Node.js installation and setup

# Prerequisites: ensure Node.js 18+ is installed
# 1) Clone the repository
# 2) Install dependencies
npm install
# 3) Build the project
npm run build

Python installation and setup

# Prerequisites: ensure Python 3.11+ is installed
# 1) Create a virtual environment
python -m venv .venv
# 2) Activate the virtual environment
source .venv/bin/activate
# 3) Install dependencies
pip install -r requirements.txt

Configure your MCP client in Claude Desktop App for a Remote Node MCP setup

{
  "mcpServers": {
    "lambda-capture-mcp": {
      "command": "npx",
      "args": [
        "mcp-remote",
        "https://mcp.lambda-capture.com/v1/mcp/",
        "--header", "Authorization: Bearer YOUR_ACCESS_TOKEN"
      ],
      "description": "RemoteMCP with Lambda Capture Macroeconomic Data API"
    }
  }
}

Additional sections

Security and environment notes: when running a local MCP server, you typically provide an API key in the environment to authorize requests. The Node-based local server uses a Node runtime and a built file path to start the MCP, while the Python-based local server runs a Python script with its own virtual environment. Make sure your API key is kept secure and not exposed in client configurations or logs.

Claude Desktop App integration: you can configure the client to connect to either the remote HTTP MCP endpoint or to a locally running MCP server by editing the claude_desktop_config.json. Examples above show how to add a remote MCP reference and how to point to a local runtime with the appropriate command and environment variable.

If you need to adjust the context window for larger queries, modify the model’s maxTokens or max_tokens settings in your client or code where applicable. This helps balance the amount of data returned by the semantic search with the model’s capacity.

Available tools

macroecon_call

Executes a macroeconomic semantic search via a remote MCP tool call, returning structured results for inflation, GDP, and other indicators.

list_tools

Lists all available MCP tools on the remote server so you can discover supported queries and actions.