mcp-ossinsight

OSSInsight MCP Server provides data analysis and insights for GitHub repositories, developers, and organizations through a set of structured tools. You can query repository analysis, developer analysis, collections, and perform natural language queries that link you directly to OSSInsight pages. The server combines OSSInsight public API access with web scraping to maximize data coverage while respecting rate limits. Use the available tools to fetch detailed metrics such as star trends, contributor activity, and project comparisons, or to browse curated collections of repositories.

Prerequisites:

• Docker installed and running on your machine (Docker Desktop for Windows/macOS or docker-engine for Linux).
• Basic familiarity with MCP (Model Context Protocol) usage and JSON-based requests.

Installation steps:

1. Ensure Docker is up and running: docker version

2. Run the OSSInsight MCP server using the provided Docker image: docker run --rm -i mcp/ossinsight

3. (Optional) If you want to customize environment variables or ports, adjust the docker run command accordingly, for example: docker run --rm -p 8080:8080 -e SOME_ENV=value -i mcp/ossinsight

4. Interact with the server using MCP requests as documented in the repository README (e.g., using the available tools like get_repo_analysis, get_developer_analysis, get_collection, list_collections, natural_language_query).

Note: The README provides an npx and a docker configuration example. If you prefer NPX, you can run the server with: npx -y @modelcontextprotocol/server-ossinsight

But for the Docker-based setup shown above, use the docker command to start the container.

Tips and notes:

• The server uses OSSInsight public API endpoints (https://api.ossinsight.io/v1/) with a fallback to web page scraping when APIs are limited.
• There is a rate limit of 600 requests per hour per IP for the OSSInsight API usage.
• If you run into connectivity issues, check your network access to api.ossinsight.io and OSSInsight web pages.
• Environment variables can be added as needed inside the Docker run command (e.g., for API keys or custom time ranges). The provided config shows an empty env map by default.
• The npm_package field is not applicable since the server is provided via Docker in this configuration.