How do I use ndpsc-case-search-mcp?

1. Click on "Install Server". 2. Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state. 3. In the chat, type @ followed by the MCP server name and your instructions, e.g., "@ndpsc-case-search-mcp Search for recent public utility cases and extract text from the latest filing" That's it! The server will respond to your query, and you can continue using it as needed. Here is a step-by-step guide with screenshots.

ndpsc-case-search-mcp

An MCP server for searching North Dakota Public Service Commission case records, dockets, and documents.

The ND PSC regulates public utilities, pipelines, mining reclamation, and other industries in North Dakota. This server provides programmatic access to their case database via the Model Context Protocol.

Tools

Tool	Description
`search_cases`	Search cases by jurisdiction, year, status, category, entity, date range, and more. Automatically splits oversized searches so broad queries can return full result sets.
`compare_case_type_by_year`	Summarize a case type across years with filed, closed, open, closure-rate, and median-days-to-close metrics.
`get_case_detail`	Get full detail for a specific case including docket count.
`get_docket_detail`	List files within a docket with direct PDF download links.
`get_document_text`	Download a PDF and extract its text content.
`get_document_pdf`	Download a PDF and return it as base64-encoded bytes.

Usage

Claude Web (Remote)

Add as a remote MCP server in Claude settings:

https://hardmcp.com/ndpsc-case-search/mcp

Claude Code (Local)

# Run the server
uv run ndpsc-case-search-mcp

# Add to Claude Code
claude mcp add ndpsc-case-search --transport http http://localhost:8000/mcp

Docker

docker build -t ndpsc-case-search-mcp .
docker run -p 8000:8000 ndpsc-case-search-mcp

Development

Requires Python 3.11+.

uv sync
uv run ndpsc-case-search-mcp

The server starts on http://0.0.0.0:8000 with a Streamable HTTP MCP endpoint at /mcp and a health check at /health.

Data Source

This server scrapes apps.psc.nd.gov/cases -- no official API exists. The PSC site pages search results at 100 rows per request; this server automatically splits broad case searches into smaller filed-date ranges so search_cases can return the full result set.