npm install -g @anthropic-ai/claude-code
npm error code 1
npm error path C:\Users\XXX\AppData\Roaming\fnm\node-versions\v22.14.0\installation\node_modules\@anthropic-ai\claude-code
npm error command failed
npm error command C:\WINDOWS\system32\cmd.exe /d /s /c node scripts/preinstall.js
npm error Error: Claude Code is not supported on Windows.
npm error Claude Code requires macOS or Linux to run properly.
npm error If you are using WSL (Windows Subsystem for Linux):
npm error 1. Make sure you are running npm install from within the WSL terminal, not from PowerShell or CMD
npm error 2. If you're still seeing this message in WSL, your environment may be incorrectly reporting as Windows
npm error Please visit https://docs.anthropic.com/en/docs/agents-and-tools/claude-code/overview#check-system-requirements for troubleshooting information.
> このコードベースを説明してください
⎿ Credit balance too low · Add funds: https://console.anthropic.com/settings/billing
> /init is analyzing your codebase…
⎿ Credit balance too low · Add funds: https://console.anthropic.com/settings/billing
ということで、カネがないと。 ここで、チャット側で課金していてもAPIは別だということを知ります。
しょうがないので、$5課金してもう一度試してみたところ
● Task(Please analyze this codebase and explain what it does. Look at the main files and structure to understand the project's …
purpose and architecture. Focus on:…)
⎿ Done (7 tool uses · 12.6k tokens · 28.5s)
● これはXXXX用のElectron+Next.jsデスクトップアプリケーションです。SQLiteデータベースを使用し、データの表示・分析機能を提供し
ます。TypeScriptで記述され、TailwindCSSでスタイリングされています。メイン機能はSQLクエリ実行、データ可視化(グラフ)、結果表示
などです。Electronのプリロードスクリプトを通じてUIとデータベース操作を安全に連携させています。
MCP Client としてはVSCodeやClineなどのツールが存在し、MCP Serverと連携。 MCP Server はLLMに対して MCP Protocolを用いて連携し、必要に応じてデータソース等から情報を提供する、、、と。 上記図の”Local Data Source”としては「Your computer’s files, databases, and services that MCP servers can securely access」とあるので、ファイルであるかもしれないし、システムであるかもしれないという位置づけ。
MCP Server が提供するもの
LLMに対して情報を提供する入口となる MCP Server ですが、これが何を提供するのかというと、下記3点となるそうです
Resources: File-like data that can be read by clients (like API responses or file contents)
Tools: Functions that can be called by the LLM (with user approval)
Prompts: Pre-written templates that help users accomplish specific tasks