data-validator-mcp

antonispaterakis/data-validator-mcp

★ 0 stars Python AI/LLM Updated 8d ago

MCP server that validates ML dataset label quality using embeddings + HiPart clustering + a local Ollama LLM judge

View on GitHub → 🔍 Audit Wallet Slippage →

Quick Install

Copy the config for your editor. Some servers may need additional setup — check the README.

Add to claude_desktop_config.json:

{
  "mcpServers": {
    "data-validator-mcp": {
      "command": "uvx",
      "args": [
        "data-validator-mcp"
      ]
    }
  }
}

Or install with pip: pip install data-validator-mcp

README Excerpt

An MCP (Model Context Protocol) server that validates the label quality of tabular datasets before they are used for ML model training. A practical adaptation of the topic-modeling pipeline described in the paper below. --- This project's lineage traces back to: > **Theocharopoulos, A., Anagnostou, A., Georgakopoulos, S., Tasoulis, S., & Plagianakos, V.**

Tools (5)

agreement_thresholdk_neighborsllm_modelmock_llmweighted_agreement

data-validator-mcp

Quick Install

README Excerpt

Tools (5)

Related AI/LLM Servers