nsip

nsip is a Rust-based MCP server that exposes the NSIP ( Sheep Genetic Evaluation Tools ) functionality via the Model Context Protocol. It includes an MCP interface with 13 tools to search animals, retrieve detailed EBV data, lineage and progeny, and generate comprehensive profiles. The server also surfaces breed groups, trait ranges, and various mating and ranking utilities to assist in flock planning. To use it, run the MCP server (for example via a Docker image) and connect to it using an MCP-enabled AI assistant or client that speaks the 13-tool protocol. Once running, you can invoke tools like search, details, lineage, progeny, profile, and compare to retrieve structured data from the NSIP database and use those results to drive AI-assisted decisions. The MCP layer is designed to work with standard MCP clients, so you can integrate with AI assistants or custom tooling that wants to reason over sheep genetic data in a contextually aware manner.

Prerequisites:

• Docker (recommended for running the MCP server from a container)
• Optional: access to the NSIP-compatible MCP image if you prefer a pre-built container

Install and run via Docker:

1. Install Docker:

   • macOS: https://docs.docker.com/desktop/mac/
   • Windows: https://docs.docker.com/desktop/windows/
   • Linux: follow your distribution’s guide

2. Pull the NSIP MCP image (or build your own if you have a custom image):

# Using a pre-built image from a registry
docker pull zircote/nsip:latest

3. Run the MCP server using Docker (exposing the MCP interface on a port of your choice):

# Run with stdout/stderr attached
docker run -i --rm -p 8080:8080 zircote/nsip:latest

4. If you need to customize environment variables (e.g., logging or port), pass them at runtime:

docker run -i --rm -p 8080:8080 -e NSIP_MCP_PORT=8080 -e NSIP_MCP_LOG_LEVEL=info zircote/nsip:latest

5. Connect your MCP client or AI assistant to the running server using the configured port and protocol (MCP over stdio or TCP, depending on your setup). Refer to the 13 MCP tools exposed by nsip for available operations.

Notes and tips:

• The MCP integration exposes 13 tools: search, details, lineage, progeny, profile, breed_groups, trait_ranges, compare, rank, inbreeding_check, mating_recommendations, flock_summary, and database_status. Each tool accepts parameters appropriate to its function (e.g., breed filters, status, page sizes, lpn_ids).
• If running via Docker, you can map a host port to the container port to enable easy access from your MCP client.
• Ensure you have access to the NSIP data sources and that any external services (e.g., databases) are reachable from the MCP server container.
• Use NSIP_MCP_LOG_LEVEL to adjust logging verbosity for troubleshooting.
• If you need to run locally for development, consider building a local image tag and using docker run with that tag to avoid pulling repeatedly.