MCP
by shaharco99
Overview
Serve as a local Model Context Protocol (MCP) server, exposing DevOps tools like kubectl, docker, and helm, with optional LLM integration for enhanced capabilities.
Installation
python server.pyEnvironment Variables
- LLM_PROVIDER
- OLLAMA_MODEL
- OPENAI_API_KEY
- OPENAI_MODEL
- GOOGLE_API_KEY
- GOOGLE_MODEL
- ANTHROPIC_API_KEY
- ANTHROPIC_MODEL
- PYTHONIMAGEVERSION
- CONTAINERNAME
- KUBECONFIG
- PYTHONUNBUFFERED
- PYTHONDONTWRITEBYTECODE
- HOME
- USERPROFILE
Security Notes
The `load_kube` utility in `server.py` explicitly disables SSL verification for Kubernetes connections when running within Docker, which is a critical security risk as it allows for man-in-the-middle attacks. While `kubectl` and `run_shell` tools implement whitelisting and checks against shell injection, the fundamental network configuration flaw compromises overall security. API keys are managed via `.env` files, which requires careful handling in shared environments.
Similar Servers
github-mcp-server
Connects AI tools (agents, assistants, chatbots) directly to GitHub's platform, enabling natural language interactions for repository management, issue/PR automation, code analysis, and workflow automation.
wcgw
An MCP server that empowers AI chat applications to execute shell commands, edit code, and manage project context on a local machine for development tasks.
code-assistant
An AI coding assistant that integrates with code editors to provide autonomous code analysis and modification, supporting multiple UI modes and the Agent Client Protocol (ACP).
responsible-vibe-mcp
Transforms any AI coding agent into a structured development partner by guiding it through battle-tested engineering workflows with persistent context and documentation.