Gcore MCP Server is an MCP (Model Context Protocol) server that enables LLM assistants to interact with Gcore Cloud services via a unified toolset. It provides access to Gcore platform resources including CDN, GPU Cloud & AI Inference, Video Streaming, WAAP, and standard cloud resources such as instances and networks. The server exposes a configurable interface through the GCORE_TOOLS environment variable, allowing a mix of predefined toolsets (management, instances, volumes, networks, baremetal, gpu_baremetal, ai_ml, billing, cleanup, list) and custom patterns (cloud.*, waap.*). Configuration modes include Default, Toolset, Pattern, and Combined, with toolsets taking precedence in combined mode and duplicates removed while preserving order. Examples show how to tailor tool selections and how to run via uvx for both temporary and persistent installation. Optional environment variables (GCORE_BASE_URL, GCORE_CLOUD_PROJECT_ID, GCORE_CLOUD_REGION_ID, GCORE_CLIENT_ID) are supported. The README also covers local development, testing with the MCP Inspector, and guidance for debugging authentication issues, all aimed at helping developers build LLM assistants that manage Gcore resources programmatically.