Managelm
AI Linux & Windows server management with Claude MCP and fully private agents using local AI
ManageLM offers private LLM support, allowing task interpretation to run entirely within your network using your own Ollama instance, and also provides a hosted model option; if AI is not desired, a No LLM mode disables all model-backed features. ManageLM enforces security through four layers, including skill scope, command allowlisting, destructive guards, and a kernel sandbox, ensuring the AI cannot execute unapproved commands.
Summary drafted from the project's own website. Every sentence is backed by text on that page and was reviewed before publishing.