Resman FAQs

Question 1

What is Resman and what problem does it solve?

Accepted Answer

Resman is a Model Context Protocol (MCP) server that exposes Kubernetes resource management as structured tools for Large Language Model (LLM) clients. It wraps standard `kubectl` commands, parses their output into fully typed Pydantic models, and returns structured JSON, making LLM interaction with Kubernetes reliable and efficient.

Question 2

Which Kubernetes resources can I manage using Resman?

Accepted Answer

Resman provides extensive tools for managing core Kubernetes resources. You can list, inspect, and switch kubeconfig contexts; list and inspect namespaces; perform pod operations like listing, getting details, fetching logs, and describing; and inspect deployments and services.

Question 3

How does Resman integrate with LLM clients like Claude Code?

Accepted Answer

Resman integrates as a local MCP server. By adding its configuration to your LLM client's `.mcp.json` file, it exposes its Kubernetes functionalities as discoverable and executable tools, allowing the LLM to interact with your cluster directly through structured commands.

Question 4

What is the benefit of 'fully typed output' in Resman?

Accepted Answer

The 'fully typed output' feature means all `kubectl` command results are validated through Pydantic v2 models. This ensures that the data returned to the LLM client is always structured, predictable JSON, eliminating parsing ambiguities and improving the reliability and robustness of automated Kubernetes tasks.

Question 5

What are the core technical requirements for running Resman?

Accepted Answer

To run Resman, you need Python 3.12+ and `kubectl` installed and accessible on your system's PATH (or configured via `RESMAN_KUBECTL_PATH`). A valid `kubeconfig` file (typically `~/.kube/config`) is also required for Kubernetes cluster access. `uv` is recommended for dependency management.

Resman

Resman

主な機能

ユースケース

主な機能

ユースケース