targetly-marketplace

Contains dockerfiles for all the marketplace images

Installation

Run this command in your terminal to add the MCP server to Claude Code.

Run in terminal:

Command

claude mcp add --transport stdio targetly-labs-targetly-marketplace node path/to/server.js \
  --env PORT="3000 (the port the MCP server will listen on)" \
  --env ADAPTER_PATH="(optional) path to targetly adapter if not the default inside the container)"

How to use

The Targetly MCP server for the Targetly Marketplace exposes a context-aware API that allows clients to fetch and reason about market data using the Model Context Protocol. The built image includes a marketplace-specific server and an adapter that bridges HTTP requests or Server-Sent Events to the MCP runtime, enabling web access to the underlying model contexts.

Once the server is running, you can interact with it through the adapter endpoints exposed inside the container. The adapter translates standard HTTP/SSE calls into MCP-context interactions, so you can request context snippets, run inference, and subscribe to streaming data as supported by the MCP server. Use the provided endpoints to discover available contexts, request updates, and integrate marketplace data with your applications via a consistent MCP interface.

How to install

Prerequisites:

Docker installed and running
Optional: Git for cloning or pulling repository contents

Clone or download the repository that contains the Official Targetly MCP Server Images.
Build the Node-based or Python-based images as shown in the repository examples:

Build Node-based Memory server image (example from README)

docker build -f servers/memory/Dockerfile -t targetly/mcp-memory:latest .

Build Python-based Time server image (example from README)

docker build -f servers/time/Dockerfile -t targetly/mcp-time:latest .

Run the built image locally to start the MCP server:

Example for a Node-based server (adjust image name as needed)

docker run -p 3000:3000 targetly/mcp-memory:latest

Example for a Python-based server (adjust image name as needed)

docker run -p 3001:3000 targetly/mcp-time:latest

Verify the server is running by curling the endpoint or visiting the adapter URL exposed by the container.
Configure your client to point to the MCP server using the provided mcp_config structure (see above) and the adapter path inside the container.

Additional notes

Tips:

The exact server name in the mcp_config should match the running container/service name (e.g., targetly-marketplace).
If your adapter path differs, update the PATH in the mcp_config's args to reflect the actual location inside the container.
Ensure environment variables like PORT are aligned with your runtime (3000–in-container port is common, but external ports may be mapped differently).
Some MCP servers require additional environment variables (e.g., authentication, database connections). Add them under the env object as placeholders or actual values as needed.
When troubleshooting, check container logs for adapter initialization errors and verify that the adapter script (targetly-adapter.js or .py) is present in the container and executable.