confluence -by-cdata

This MCP server provides a read-only interface to your Confluence data via the CData JDBC Driver. It wraps the JDBC driver and exposes data through a lightweight MCP interface, allowing LLMs to query live Confluence content without writing SQL. The server is designed to run locally and communicates over stdio, meaning it is intended for use with clients on the same machine (e.g., Claude Desktop or other MCP clients). Tools exposed follow a naming convention based on your configured server name (e.g., confluence_by_cdata_get_tables, confluence_by_cdata_get_columns, confluence_by_cdata_run_query). The available tools allow discovering tables, inspecting columns, and executing SQL SELECT queries to retrieve data in CSV format. To use it, configure Claude Desktop (or your chosen MCP client) with a CLAUDE-compatible config file referencing the Java process and the PRP file that contains your JDBC connection details. The JSON-RPC examples show how to call these tools programmatically, returning results that can be parsed and surfaced by the AI client.

Prerequisites:

• Java Runtime Environment (JRE) or JDK installed
• Maven (for building the MCP server)
• The CData JDBC Driver for Confluence (license as needed)

Install steps:

1. Clone the repository and navigate into it:

   git clone https://github.com/cdatasoftware/confluence-mcp-server-by-cdata.git
   cd confluence-mcp-server-by-cdata
   

2. Build the MCP server JAR with dependencies:

   mvn clean install
   

   This generates a file named something like CDataMCP-jar-with-dependencies.jar in the target directory.
3. Download and install the CData JDBC Driver for Confluence from: https://www.cdata.com/drivers/confluence/download/jdbc
4. License the JDBC Driver (example steps):
   • Locate the lib folder of the JDBC driver installation
   • Run: java -jar cdata.jdbc.confluence.jar --license
   • Enter your name, email, and a license key or TRIAL as appropriate.
5. Create a PRP (properties) file for your connection with the required fields (Prefix, ServerName, ServerVersion, DriverPath, DriverClass, JdbcUrl, Tables).
6. Run the MCP server:

   java -jar CDataMCP-jar-with-dependencies.jar /PATH/TO/confluence.prp
   

   The server will start and listen via stdio for MCP client connections.

Notes and tips:

• The MCP server is read-only at this level; to enable full read/write capabilities you would need the full MCP server setup with writable access.
• The server uses stdio for communication; ensure your MCP client launches on the same machine or is configured to pipe stdio appropriately.
• The configuration file path in the command should point to your prepared confluence.prp with the correct DriverPath, JdbcUrl, and other properties set.
• If Claude Desktop or another client cannot see the MCP server, fully quit and restart the client to refresh the MCP server list.
• If you need to switch Confluence instances, update the confluence.prp and restart the server so the new JDBC connection is used.
• You can discover tables with the {servername}_get_tables tool and inspect columns with {servername}_get_columns; then query data with {servername}_run_query.