Dataverse MCP Server

Example Use Cases

• 1

  Fetch structured dataset metadata from Dataverse by DOI (via Croissant/Schema.org) to answer user questions or populate context.

• 2

  Multilingual dataset descriptions and file inventories to support search, summarization, or verification workflows.

• 3

  Integrate Dataverse dataset lookups into editor or chat-based workflows (Cursor, VS Code, Claude Desktop, Windsurf, Zed) using an SSE MCP endpoint.

Description

This MCP server provides programmatic access to Dataverse datasets using the Croissant export (Schema.org-based) and exposes tool endpoints (e.g., get_croissant_record) over an SSE interface. It is intended for use by AI agents and editor integrations (Cursor, VS Code, Claude Desktop, Windsurf, Zed) to fetch dataset metadata by DOI, support multilingual metadata, and surface file-level details. The README documents a public instance at https://mcp.dataverse.org and includes instructions to run a local instance via docker-compose for development or private use.