datascript

Datascript MCP Server provides a Clojure-based interface for querying DataScript datasets via JSON-RPC, with capabilities for graph path search, visualization integration, and entity/property exploration. The server runs in the current process and communicates over STDIO using the MCP protocol. Typical usage starts the server with the Clojure CLI, then you can send JSON-RPC requests (e.g., initialize, followed by tool discovery and method invocations) to explore data, relationships, and derived insights. The README highlights several core capabilities: using graph path search to visualize relationships between entities, querying parent-child relationships, loading example datasets and computing aggregates (totals), and querying individual entity properties. You can pair this server with the Visualization MCP Server to render relationship graphs and with clients like Claude Desktop for scripted workflows. The sample health check shows you can issue an initialize request and receive a response that includes available tools (such as listChanged) and server information, confirming that the server is ready to process subsequent RPC calls.

Prerequisites:

• Java Runtime Environment (JRE) or JDK installed
• Clojure CLI installed (clojure) and available on PATH
• Access to the repository containing the Datascript MCP Server

Installation steps:

1. Install prerequisites

   • Install Java from https://adoptium.net/
   • Install Clojure CLI from https://clojure.org/guides/getting_started

2. Clone the repository

   • git clone <repository-url-for-datascript-mcp-server>
   • cd datascript-mcp-server

3. Install dependencies (if your project uses deps.edn or similar)

   • Ensure clojure is available in PATH and the project can resolve dependencies via the MCP profile

4. Run the MCP server

   • clojure -X:mcp
   • Alternatively, start an interactive REPL and evaluate (mcp-server)

5. Verify the server is running

   • Send an initialize JSON-RPC request over STDIO as described in the README to confirm responses and available tools.

Tips and notes:

• The server communicates via JSON-RPC over STDIO. For tooling integrations, you can script requests to initialize and list available tools (e.g., listChanged).
• The readme suggests integration with the Visualization MCP Server for graph visualizations; consider using the combined workflow to explore relationships in DataScript datasets.
• If deploying with Claude Desktop or other clients, you can wrap the clojure -X:mcp invocation in a shell command, as shown in the example client configuration.
• If you plan to load external datasets or enable additional features, check for environment-specific configuration options or paths under src/datascript_mcp and related modules.
• Troubleshooting: ensure CLJ/Korma dependencies are resolved, verify Java version compatibility, and confirm that the MCP mode (-X:mcp) is enabled in your clojure config.