Dingo MCP Server

Example Use Cases

• 1

  Evaluate dataset quality automatically using a mix of heuristic rules and LLM-based prompts across text and multimodal data.

• 2

  An MCP-compatible endpoint to integrate Dingo's evaluation workflows into retrieval-augmented or agent-driven pipelines (e.g., Cursor integration).

• 3

  Run large-scale evaluations with local or Spark execution and generate traceable summary and detailed GUI reports for dataset issues.

Description

Dingo MCP Server is part of the Dingo data-quality evaluation toolkit and provides an experimental MCP interface to run rule-based and LLM-based dataset quality checks. It enables integration with clients like Cursor and supports local CLI/SDK, Spark execution, LLM integrations (OpenAI, local models), and GUI visualization of results. The server is intended to help automate and standardize dataset assessment workflows, produce traceable reports, and enable agent-driven evaluation and RAG/hallucination checks.