eka_mcp_server
Eka MCP Server
claude mcp add --transport stdio eka-care-eka_mcp_server uvx eka_mcp_server --eka-api-host https://api.eka.care --client-id <client_id> --client-secret <client_secret>
How to use
The Eka MCP Server exposes two primary knowledge bases tailored for the Indian healthcare context: Indian Branded Drug Search and Indian Treatment Protocol Search. The Indian Branded Drug Search lets you look up branded drugs available in India and returns metadata such as generic composition and manufacturer details to help ground AI responses. The Indian Treatment Protocol Search provides access to standardized guidelines and treatment pathways from Indian authorities (e.g., ICMR, RSSDI), covering hundreds of conditions and protocols. You can query these tools to retrieve structured drug information or treatment pathways to support clinical decision making and to improve the factual grounding of AI-assisted conversations. When running with a client ID and secret, the server authenticates to Eka’s API host and exposes endpoints that the MCP client (e.g., Claude or a chat interface) can invoke via the MCP protocol.
To use the server with debugging or integration tooling, you can launch the MCP Inspector, which connects to the running MCP server via the same command you use in production and provides a browser-based debugging surface. In practice, you run the server through uvx and point your MCP client at https://api.eka.care, supplying your client credentials. The tools are designed to be invoked by LLMs or orchestration components that understand the MCP tool invocation pattern described in the repository.
How to install
Prerequisites:
- A supported runtime (Python) and the UV runtime for running MCP servers locally (via uvx).
- A Eka Care developer account to obtain a client_id and client_secret.
Installation steps:
- Install UV (if not already installed):
- curl -LsSf https://astral.sh/uv/install.sh | sh
- Install the Eka MCP Server package from PyPI so uvx can execute it:
- pip install eka_mcp_server
- Register or confirm your Eka API credentials:
- Obtain client_id and client_secret from https://console.eka.care
- Run the MCP server via UVX:
- uvx eka_mcp_server --eka-api-host https://api.eka.care --client-id <client_id> --client-secret <client_secret>
- Verify you can connect with an MCP client (e.g., MCP Inspector) using the same host and credentials.
Configuration example (config can be placed in your MCP client config):
{
"mcpServers": {
"eka-mcp-server": {
"command": "uvx",
"args": [
"eka_mcp_server",
"--eka-api-host",
"https://api.eka.care",
"--client-id",
"<client_id>",
"--client-secret",
"<client_secret>"
]
}
}
}
Additional notes
Tips and common considerations:
- The MCP server runs over stdio; debugging can be challenging. Use the MCP Inspector for a better debugging experience.
- Ensure the uvx path is correct if uvx isn’t discoverable in your shell (you may need to provide the full path to the uvx executable).
- If you encounter spawn uvx ENOENT, install UV and verify the full path to uvx in your config as shown in the Troubleshooting section of the README.
- Keep your client_id and client_secret secure. Do not commit them to source control.
- The server relies on authenticated access to Eka’s API; credentials are required even for local testing.
Related MCP Servers
mcp-vegalite
MCP server from isaacwasserman/mcp-vegalite-server
github-chat
A Model Context Protocol (MCP) for analyzing and querying GitHub repositories using the GitHub Chat API.
nautex
MCP server for guiding Coding Agents via end-to-end requirements to implementation plan pipeline
pagerduty
PagerDuty's official local MCP (Model Context Protocol) server which provides tools to interact with your PagerDuty account directly from your MCP-enabled client.
futu-stock
mcp server for futuniuniu stock
mcp -boilerplate
Boilerplate using one of the 'better' ways to build MCP Servers. Written using FastMCP