Orm Discovery

This MCP server exposes HTTP endpoints to interact with the ORM discovery workflow. You can send natural language queries, fetch message history, and invoke discovery tools through simple HTTP requests. It is designed for easy integration with client tools and automation pipelines that drive data discovery tasks from your applications and workflows.

How to use

To start interacting with the MCP server, first connect to its message stream and then issue requests to the available endpoints. The server provides an HTTP interface for querying, retrieving history, and calling tools.

Start the real-time connection to the server’s event stream with the client. This establishes a live channel you can monitor for updates as actions are performed.

uv run client.py http://localhost:8080/sse --api

Send a query to the server to process a natural language request. The server expects a JSON payload describing the text you want analyzed or acted upon.

curl -X POST http://localhost:8000/query -H "Content-Type: application/json" -d \
'{"text":"Find the top authors on oreilly who write about ai. List their names and the titles of their top books along with links."}'

Retrieve the history of messages to review prior interactions with the server.

curl http://localhost:8000/messages

You can call a specific tool exposed by the MCP server to perform a targeted action, such as content search. This demonstrates how to invoke a named tool with arguments.

curl -X POST http://0.0.0.0:8000/call-tool \
   -H "Content-Type: application/json" \
   -d '{
      "tool_name": "search_content",
      "tool_args": { "query": "python" }}'

Available tools

search_content

Tool to search content with a query, enabling discovery of relevant results from a data source.