Bullhorn CRM MCP Server by CData

What is Bullhorn CRM MCP Server by CData?

The Bullhorn CRM MCP Server by CData is a local, read-only MCP server that allows large language models (LLMs) like Claude Desktop to query live data from Bullhorn CRM. It wraps CData’s JDBC Driver for Bullhorn CRM, exposing that data as relational SQL models and providing a simple MCP interface for natural-language querying — no SQL required.

How to use Bullhorn CRM MCP Server by CData?

1. Clone the repository and build with mvn clean install to produce a JAR.
2. Download, install, and license the CData JDBC Driver for Bullhorn CRM (use java -jar cdata.jdbc.bullhorncrm.jar --license).
3. Configure a connection string using the driver’s utility, then create a .prp file with JDBC properties.
4. Add the server to claude_desktop_config.json with the path to Java, the JAR, and the .prp file.
5. Run or restart Claude Desktop to see the tools. The server uses stdio, so it must run on the same machine as the client.

Key features of Bullhorn CRM MCP Server by CData

• Read-only access to live Bullhorn CRM data via MCP tools
• Exposes tables and columns with get_tables and get_columns
• Execute arbitrary SQL SELECT queries through run_query
• Integrates with LLM clients for natural-language data questions
• Built on the CData JDBC Driver for Bullhorn CRM

Use cases of Bullhorn CRM MCP Server by CData

• Ask an LLM to summarize correlations between closed‑won opportunities and account industries
• Retrieve the number of open tickets in a specific project
• Query today’s calendar events from Bullhorn CRM using natural language
• Explore Bullhorn CRM data schema without writing SQL queries

FAQ from Bullhorn CRM MCP Server by CData

Is this server read-only?

Yes. This project builds a local, read‑only MCP server. For full read, write, update, delete, and action capabilities, CData offers a separate MCP Server for Bullhorn CRM.

What are the dependencies and runtime requirements?

You need Java, Maven (to build), and the CData JDBC Driver for Bullhorn CRM (installed and licensed). The server runs locally via stdio.

How does authentication work?

If the data source uses OAuth (as Bullhorn CRM does), you authenticate in your browser when testing the connection string. The JDBC driver handles token refresh automatically via InitiateOAuth=GETANDREFRESH.

Where does data live and what transport is used?

Data is queried live from Bullhorn CRM; no data is stored locally by the server. The server communicates with the client via stdio, so it must run on the same machine.

What are the known limits?

The server is read‑only and supports only SELECT queries. It cannot write, update, or delete data. It also requires local installation and cannot be accessed remotely.